技术领域Technical field
本申请涉及计算机技术领域，尤其涉及一种商品推荐文本生成方法、装置及设备。The present application relates to the field of computer technology, and in particular to a method, device and equipment for generating product recommendation text.
背景技术Background technique
随着互联网技术的不断发展，线上购物逐渐成为一种主流的购物方式，极大地方便了用户的生活。由于商品种类和数量繁多，为了提高浏览量以及购买量，电商类应用程序(Application，APP)在显示商品时，通常会显示商品推荐理由等文本。With the continuous development of Internet technology, online shopping has gradually become a mainstream shopping method, which greatly facilitates users' lives. Due to the wide variety and quantity of products, in order to increase the number of views and purchases, e-commerce applications (Applications, APPs) usually display text such as reasons for product recommendations when displaying products.
在相关技术中，各个商品的商品推荐文本通常需要运营人员人工进行生成和投放显示，同时运营人员还需要人工对文案进行频繁的动态更新。这种商品推荐文本的配置方式，人力资源耗费较大，时间耗费较多，并且人工生成的商品推荐文本准确性也不高，无法契合消费者的实际需求。In related technologies, product recommendation text for each product usually requires operators to manually generate, place and display the text. At the same time, operators also need to manually update the text frequently and dynamically. This method of configuring product recommendation text consumes a lot of human resources and time, and the accuracy of the manually generated product recommendation text is not high and cannot meet the actual needs of consumers.
发明内容Contents of the invention
本申请的多个方面提供一种商品推荐文本生成方法、装置及设备，用以实现商品推荐文案的自动生成，能够减少人力资源的耗费，节省时间成本，同时也能够提高商品推荐文本的准确性，契合消费者的实际需求。Various aspects of this application provide a product recommendation text generation method, device and equipment to realize the automatic generation of product recommendation copywriting, which can reduce the consumption of human resources, save time and cost, and also improve the accuracy of product recommendation text. , in line with the actual needs of consumers.
第一方面，本申请实施例提供一种商品推荐文本生成方法，包括：In the first aspect, embodiments of the present application provide a method for generating product recommendation text, including:
获取第一关联数据、第二关联数据以及目标格式文本；所述第一关联数据包括商品标识以及商品标签；所述第二关联数据包括用户标识以及人群标签；所述目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；Obtain the first associated data, the second associated data and the target format text; the first associated data includes product identification and product tags; the second associated data includes user identification and group tags; the target format text includes a preset description At least two of text, product tags, time period and behavioral data;
根据所述第二关联数据，创建所述人群标签对应的人群标识并确定所述人群标识包括的用户标识；According to the second associated data, create a crowd identification corresponding to the crowd tag and determine the user identification included in the crowd identification;
在所述目标格式文本中包括所述行为数据的情况下，获取所述商品标识在时间周期内对应的历史数据，并根据所述历史数据、人群标识以及所述行为数据，确定所述人群标识对应的目标行为数据；When the target format text includes the behavioral data, obtain the historical data corresponding to the product identification within the time period, and determine the crowd identification based on the historical data, the crowd identification and the behavioral data Corresponding target behavior data;
基于所述目标行为数据，按照所述目标格式文本生成所述商品标识对应的商品推荐文本。Based on the target behavior data, product recommendation text corresponding to the product identification is generated according to the target format text.
在一种可能的实施方式中，所述创建所述人群标签对应的人群标识并确定所述人群标识包括的用户标识，包括：In a possible implementation, creating a crowd identification corresponding to the crowd tag and determining a user ID included in the crowd identification includes:
读取所述第二关联数据中的多个人群标签，分别创建各个人群标签对应的人群标识；Read multiple crowd tags in the second associated data, and create crowd identifiers corresponding to each crowd tag;
根据所述第二关联数据中用户标识与人群标签的关联关系，确定各个所述人群标识包括的用户标识；Determine the user identifiers included in each of the crowd identifiers according to the association relationship between the user identifiers and the crowd tags in the second associated data;
存储所述人群标识以及所述人群标识包括的用户标识。The crowd identification and the user identification included in the crowd identification are stored.
在一种可能的实施方式中，所述获取所述商品标识在时间周期内对应的历史数据，包括：In a possible implementation, the obtaining historical data corresponding to the commodity identification within a time period includes:
周期性检测所述第一关联数据与所述第二关联数据的更新状态；Periodically detect the update status of the first associated data and the second associated data;
在所述第一关联数据与所述第二关联数据更新完成后，获取所述商品标识在时间周期内的历史数据；所述历史数据包括商品标识对应的历史行为数据。After the first associated data and the second associated data are updated, historical data of the product identification within a time period is obtained; the historical data includes historical behavior data corresponding to the product identification.
在一种可能的实施方式中，所述确定所述人群标识对应的目标行为数据，包括：In a possible implementation, determining the target behavior data corresponding to the crowd identifier includes:
确定所述行为数据对应的目标行为类型；Determine the target behavior type corresponding to the behavior data;
在所述历史数据中，确定所述人群标识对应的目标行为类型的数量，得到所述人群标识对应的目标行为数据。In the historical data, the number of target behavior types corresponding to the crowd identification is determined to obtain target behavior data corresponding to the crowd identification.
在一种可能的实施方式中，所述方法还包括：In a possible implementation, the method further includes:
基于所述商品标识、所述人群标识以及所述商品推荐文本之间的关联关系，将所述商品标识、所述人群标识以及所述商品推荐文本存储至目标数据表中；Based on the association between the product identification, the crowd identification and the product recommendation text, store the product identification, the crowd identification and the product recommendation text into a target data table;
接收客户端发送的数据导出请求，生成目标文件并向客户端发送所述目标文件。Receive the data export request sent by the client, generate a target file and send the target file to the client.
在一种可能的实施方式中，所述商品标签包括地理位置标签、季节时令标签、身份标签、类别标签其中至少一种。In a possible implementation, the product tag includes at least one of a geographical location tag, a seasonal tag, an identity tag, and a category tag.
第二方面，本申请实施例提供一种商品推荐文本生成方法，包括：In the second aspect, embodiments of the present application provide a method for generating product recommendation text, including:
响应于第一用户的第一触控操作，确定第一关联数据以及第二关联数据；所述第一关联数据包括商品标识以及商品标签；所述第二关联数据包括用户标识以及人群标签；In response to the first user's first touch operation, first associated data and second associated data are determined; the first associated data includes product identification and product tags; the second associated data includes user identification and crowd tags;
响应于第一用户的第二触控操作，确定目标格式文本；所述目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；In response to the second touch operation of the first user, determine the target format text; the target format text includes at least two of preset description text, product tags, time period and behavior data;
向服务端发送所述第一关联数据、所述第二关联数据以及所述目标格式文本，以使得所述服务端基于所述第二关联数据创建所述人群标签对应的人群标识并确定所述人群标识包括的用户标识，并且基于所述目标格式文本生成所述商品标识对应的商品推荐文本。Send the first associated data, the second associated data and the target format text to the server, so that the server creates a crowd identifier corresponding to the crowd tag based on the second associated data and determines the The crowd identification includes a user identification, and product recommendation text corresponding to the product identification is generated based on the target format text.
在一种可能的实施方式中，所述响应于第一用户的第二触控操作，确定目标格式文本，包括：In a possible implementation, determining the target format text in response to the first user's second touch operation includes:
响应于所述第二触控操作，基于所述第二触控操作对应的文本元素生成目标格式文本；或者，In response to the second touch operation, generate target format text based on the text element corresponding to the second touch operation; or,
响应于所述第二触控操作，基于所述第二触控操作对应的文本元素生成目标格式文本，并确定所述目标格式文本对应的其他历史数据以及查询条件。In response to the second touch operation, a target format text is generated based on the text element corresponding to the second touch operation, and other historical data and query conditions corresponding to the target format text are determined.
在一种可能的实施方式中，所述方法还包括：In a possible implementation, the method further includes:
响应于第一用户的数据导出操作，向服务端发送数据导出请求，以使得所述服务端基于所述数据导出请求生成目标文件；所述目标文件中包括具备关联关系的商品标识、人群标识以及商品推荐文本；In response to the first user's data export operation, a data export request is sent to the server, so that the server generates a target file based on the data export request; the target file includes an associated product identifier, a group identifier, and Product recommendation text;
接收所述服务端返回的所述目标文件。Receive the target file returned by the server.
第三方面，本申请实施例提供一种商品推荐文本生成方法，包括：In a third aspect, embodiments of the present application provide a method for generating product recommendation text, including:
响应于第二用户的页面加载操作，确定所述第二用户的用户标识；In response to the second user's page loading operation, determining the user identification of the second user;
向服务端发送所述用户标识，以使得服务端根据所述用户标识确定所述第二用户对应的人群标识以及目标商品标识，并基于所述人群标识以及目标商品标识从商品推荐文本中确定所述第二用户对应的目标商品推荐文本；Send the user identification to the server, so that the server determines the crowd identification and the target product identification corresponding to the second user based on the user identification, and determines the product recommendation text based on the crowd identification and the target product identification. The target product recommendation text corresponding to the second user;
展示所述目标商品标识对应的目标商品以及目标商品推荐文本。Display the target product corresponding to the target product identification and the target product recommendation text.
第四方面，本申请实施例提供一种商品推荐文本生成方法，包括：In the fourth aspect, embodiments of the present application provide a method for generating product recommendation text, including:
获取第二用户的用户标识，并根据所述用户标识，确定所述第二用户对应的人群标识以及目标商品标识；Obtain the user identification of the second user, and determine the group identification and target product identification corresponding to the second user based on the user identification;
基于所述人群标识以及目标商品标识，从商品推荐文本中确定所述第二用户对应的目标商品推荐文本；Based on the crowd identification and the target product identification, determine the target product recommendation text corresponding to the second user from the product recommendation text;
向客户端发送所述目标商品标识以及所述目标商品推荐文本。Send the target product identification and the target product recommendation text to the client.
第五方面，本申请实施例提供一种商品推荐文本生成装置，包括：In the fifth aspect, embodiments of the present application provide a device for generating product recommendation text, including:
第一获取模块，用于获取第一关联数据、第二关联数据以及目标格式文本；所述第一关联数据包括商品标识以及商品标签；所述第二关联数据包括用户标识以及人群标签；所述目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；The first acquisition module is used to obtain the first associated data, the second associated data and the target format text; the first associated data includes product identification and product tags; the second associated data includes user identification and group tags; The target format text includes at least two of preset description text, product tags, time period and behavioral data;
创建模块，用于根据所述第二关联数据，创建所述人群标签对应的人群标识并确定所述人群标识包括的用户标识；A creation module, configured to create a crowd identification corresponding to the crowd tag and determine the user ID included in the crowd identification according to the second associated data;
第二获取模块，用于在所述目标格式文本中包括所述行为数据的情况下，获取所述商品标识在时间周期内对应的历史数据，并根据所述历史数据、人群标识以及所述行为数据，确定所述人群标识对应的目标行为数据；The second acquisition module is used to obtain the historical data corresponding to the product identification within the time period when the behavioral data is included in the target format text, and obtain the historical data based on the historical data, the crowd identification and the behavioral data. Data to determine the target behavior data corresponding to the crowd identification;
生成模块，用于基于所述目标行为数据，按照所述目标格式文本生成所述商品标识对应的商品推荐文本。A generating module, configured to generate product recommendation text corresponding to the product identification according to the target format text based on the target behavior data.
在一种可能的实施方式中，所述创建模块，具体用于：In a possible implementation, the creation module is specifically used to:
读取所述第二关联数据中的多个人群标签，分别创建各个人群标签对应的人群标识；Read multiple crowd tags in the second associated data, and create crowd identifiers corresponding to each crowd tag;
根据所述第二关联数据中用户标识与人群标签的关联关系，确定各个所述人群标识包括的用户标识；Determine the user identifiers included in each of the crowd identifiers according to the association relationship between the user identifiers and the crowd tags in the second associated data;
存储所述人群标识以及所述人群标识包括的用户标识。The crowd identification and the user identification included in the crowd identification are stored.
在一种可能的实施方式中，所述第二获取模块，具体用于：In a possible implementation, the second acquisition module is specifically used to:
周期性检测所述第一关联数据与所述第二关联数据的更新状态；Periodically detect the update status of the first associated data and the second associated data;
在所述第一关联数据与所述第二关联数据更新完成后，获取所述商品标识在时间周期内的历史数据；所述历史数据包括商品标识对应的历史行为数据。After the first associated data and the second associated data are updated, historical data of the product identification within a time period is obtained; the historical data includes historical behavior data corresponding to the product identification.
在一种可能的实施方式中，所述第二获取模块，具体用于：In a possible implementation, the second acquisition module is specifically used to:
确定所述行为数据对应的目标行为类型；Determine the target behavior type corresponding to the behavior data;
在所述历史数据中，确定所述人群标识对应的目标行为类型的数量，得到所述人群标识对应的目标行为数据。In the historical data, the number of target behavior types corresponding to the crowd identification is determined to obtain target behavior data corresponding to the crowd identification.
在一种可能的实施方式中，所述装置还用于：In a possible implementation, the device is also used for:
基于所述商品标识、所述人群标识以及所述商品推荐文本之间的关联关系，将所述商品标识、所述人群标识以及所述商品推荐文本存储至目标数据表中；Based on the association between the product identification, the crowd identification and the product recommendation text, store the product identification, the crowd identification and the product recommendation text into a target data table;
接收客户端发送的数据导出请求，生成目标文件并向客户端发送所述目标文件。Receive the data export request sent by the client, generate a target file and send the target file to the client.
在一种可能的实施方式中，所述商品标签包括地理位置标签、季节时令标签、身份标签、类别标签其中至少一种。In a possible implementation, the product tag includes at least one of a geographical location tag, a seasonal tag, an identity tag, and a category tag.
第六方面，本申请实施例提供一种商品推荐文本生成装置，包括：In a sixth aspect, embodiments of the present application provide a device for generating product recommendation text, including:
第一确定模块，用于响应于第一用户的第一触控操作，确定第一关联数据以及第二关联数据；所述第一关联数据包括商品标识以及商品标签；所述第二关联数据包括用户标识以及人群标签；A first determination module, configured to determine first associated data and second associated data in response to the first user's first touch operation; the first associated data includes a product identification and a product tag; the second associated data includes User identification and group tags;
第二确定模块，用于响应于第一用户的第二触控操作，确定目标格式文本；所述目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；a second determination module, configured to determine target format text in response to the first user's second touch operation; the target format text includes at least two of preset description text, product tags, time periods, and behavior data;
发送模块，用于向服务端发送所述第一关联数据、所述第二关联数据以及所述目标格式文本，以使得所述服务端基于所述第二关联数据创建所述人群标签对应的人群标识并确定所述人群标识包括的用户标识，并且基于所述目标格式文本生成所述商品标识对应的商品推荐文本。A sending module, configured to send the first associated data, the second associated data and the target format text to the server, so that the server creates a crowd corresponding to the crowd tag based on the second associated data. Identify and determine user identifiers included in the crowd identifier, and generate product recommendation text corresponding to the product identifier based on the target format text.
在一种可能的实施方式中，所述第二确定模块，具体用于：In a possible implementation, the second determination module is specifically used to:
响应于所述第二触控操作，基于所述第二触控操作对应的文本元素生成目标格式文本；或者，In response to the second touch operation, generate target format text based on the text element corresponding to the second touch operation; or,
响应于所述第二触控操作，基于所述第二触控操作对应的文本元素生成目标格式文本，并确定所述目标格式文本对应的其他历史数据以及查询条件。In response to the second touch operation, a target format text is generated based on the text element corresponding to the second touch operation, and other historical data and query conditions corresponding to the target format text are determined.
在一种可能的实施方式中，所述装置还用于：In a possible implementation, the device is also used for:
响应于第一用户的数据导出操作，向服务端发送数据导出请求，以使得所述服务端基于所述数据导出请求生成目标文件；所述目标文件中包括具备关联关系的商品标识、人群标识以及商品推荐文本；In response to the first user's data export operation, a data export request is sent to the server, so that the server generates a target file based on the data export request; the target file includes an associated product identifier, a group identifier, and Product recommendation text;
接收所述服务端返回的所述目标文件。Receive the target file returned by the server.
第七方面，本申请实施例提供一种商品推荐文本生成装置，包括：In the seventh aspect, embodiments of the present application provide a device for generating product recommendation text, including:
确定模块，用于响应于第二用户的页面加载操作，确定所述第二用户的用户标识；A determining module, configured to determine the user identification of the second user in response to the second user's page loading operation;
发送模块，用于向服务端发送所述用户标识，以使得服务端根据所述用户标识确定所述第二用户对应的人群标识以及目标商品标识，并基于所述人群标识以及目标商品标识从商品推荐文本中确定所述第二用户对应的目标商品推荐文本；A sending module configured to send the user identification to the server, so that the server determines the crowd identification and target product identification corresponding to the second user based on the user identification, and extracts the product from the product based on the crowd identification and the target product identification. Determine the target product recommendation text corresponding to the second user in the recommendation text;
展示模块，用于展示所述目标商品标识对应的目标商品以及目标商品推荐文本。A display module is used to display the target product corresponding to the target product identification and the target product recommendation text.
第八方面，本申请实施例提供一种商品推荐文本生成装置，包括：In an eighth aspect, embodiments of the present application provide a device for generating product recommendation text, including:
获取模块，用于获取第二用户的用户标识，并根据所述用户标识，确定所述第二用户对应的人群标识以及目标商品标识；An acquisition module, configured to obtain the user identification of the second user, and determine the group identification and target product identification corresponding to the second user based on the user identification;
确定模块，用于基于所述人群标识以及目标商品标识，从商品推荐文本中确定所述第二用户对应的目标商品推荐文本；A determination module configured to determine the target product recommendation text corresponding to the second user from the product recommendation text based on the crowd identification and the target product identification;
发送模块，用于向客户端发送所述目标商品标识以及所述目标商品推荐文本。A sending module, configured to send the target product identification and the target product recommendation text to the client.
第九方面，本申请实施例提供一种电子设备，包括：存储器和处理器；In a ninth aspect, embodiments of the present application provide an electronic device, including: a memory and a processor;
所述存储器存储计算机执行指令；The memory stores computer execution instructions;
所述处理器执行所述存储器存储的计算机执行指令，使得所述处理器执行第一方面至第四方面任一项所述的商品推荐文本生成方法。The processor executes the computer execution instructions stored in the memory, so that the processor executes the product recommendation text generation method described in any one of the first to fourth aspects.
第十方面，本申请实施例提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机执行指令，当所述计算机执行指令被处理器执行时用于实现第一方面至第四方面任一项所述的商品推荐文本生成方法。In a tenth aspect, embodiments of the present application provide a computer-readable storage medium in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the first aspect to the third aspect. The product recommendation text generation method described in any of the four aspects.
第十一方面，本申请实施例提供一种计算机程序产品，包括计算机程序，该计算机程序被处理器执行时实现第一方面至第四方面任一项所示的商品推荐文本生成方法。In an eleventh aspect, embodiments of the present application provide a computer program product, including a computer program that, when executed by a processor, implements the method for generating product recommendation text shown in any one of the first to fourth aspects.
在本申请实施例中，服务端获取第一关联数据、第二关联数据以及目标格式文本；第一关联数据包括商品标识以及商品标签；第二关联数据包括用户标识以及人群标签；目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；根据第二关联数据，创建人群标签对应的人群标识并确定人群标识包括的用户标识；在目标格式文本中包括行为数据的情况下，获取商品标识在时间周期内对应的历史数据，并根据历史数据、人群标识以及行为数据，确定人群标识对应的目标行为数据；基于目标行为数据，按照目标格式文本生成商品标识对应的商品推荐文本。本申请实施例中，服务端可以根据运营人员配置的目标格式文本来自动统计目标行为数据并自动生成、更新商品推荐文本，无需运营人员进行文案的人工生成，能够降低人力资源的耗费，节省时间成本，同时服务端可以根据运营人员上传的第二关联数据来自动创建人群标识，并且将人群标识与多个用户标识相关联，这样能够快速确定消费者的所属的人群标识，结合多样化的商品标签，可以提高文案推荐文本的准确性，更加契合消费者的实际需求，能够增强对消费者的吸引力。In the embodiment of this application, the server obtains the first associated data, the second associated data and the target format text; the first associated data includes product identification and product tags; the second associated data includes user identification and group tags; the target format text includes Preset at least two of description text, product tag, time period and behavioral data; create a crowd identifier corresponding to the crowd tag according to the second associated data and determine the user ID included in the crowd identifier; include behavioral data in the target format text Next, obtain the historical data corresponding to the product identification within the time period, and determine the target behavior data corresponding to the crowd identification based on historical data, crowd identification, and behavioral data; based on the target behavior data, generate product recommendations corresponding to the product identification according to the target format text text. In the embodiment of this application, the server can automatically count target behavior data and automatically generate and update product recommendation text based on the target format text configured by the operator. There is no need for the operator to manually generate copywriting, which can reduce the consumption of human resources and save time. cost, and at the same time, the server can automatically create a crowd ID based on the second associated data uploaded by the operator, and associate the crowd ID with multiple user IDs, so that the consumer's crowd ID can be quickly determined and combined with diversified products Tags can improve the accuracy of copywriting recommendation text, better meet the actual needs of consumers, and enhance the appeal to consumers.
附图说明Description of the drawings
此处所说明的附图用来提供对本申请的进一步理解，构成本申请的一部分，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：The drawings described here are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application. In the attached picture:
图1为本申请示例性实施例提供的一种应用场景的示意图；Figure 1 is a schematic diagram of an application scenario provided by an exemplary embodiment of the present application;
图2为本申请示例性实施例提供的一种商品推荐文本生成方法的流程示意图；Figure 2 is a schematic flowchart of a method for generating product recommendation text provided by an exemplary embodiment of the present application;
图3为本申请示例性实施例提供的另一种商品推荐文本生成方法的流程示意图；Figure 3 is a schematic flow chart of another product recommendation text generation method provided by an exemplary embodiment of the present application;
图4为本申请示例性实施例提供的一种服务端商品推荐文本生成过程的技术链路图；Figure 4 is a technical link diagram of a server-side product recommendation text generation process provided by an exemplary embodiment of the present application;
图5为本申请示例性实施例提供的另一种商品推荐文本生成方法的流程示意图；Figure 5 is a schematic flow chart of another product recommendation text generation method provided by an exemplary embodiment of the present application;
图6为本申请示例性实施例提供的一种商品推荐文本配置页面的示意图；Figure 6 is a schematic diagram of a product recommendation text configuration page provided by an exemplary embodiment of the present application;
图7为本申请示例性实施例提供的一种商品推荐文本导出页面的示意图；Figure 7 is a schematic diagram of a product recommendation text export page provided by an exemplary embodiment of the present application;
图8为本申请示例性实施例提供的一种商品推荐文本第一用户配置的技术链路示意图；Figure 8 is a schematic diagram of technical links configured by a first user of product recommendation text provided by an exemplary embodiment of the present application;
图9为本申请示例性实施例提供的另一种商品推荐文本生成方法的流程示意图；Figure 9 is a schematic flow chart of another product recommendation text generation method provided by an exemplary embodiment of the present application;
图10为本申请示例性实施例提供的一种商品推荐文本显示页面的示意图；Figure 10 is a schematic diagram of a product recommendation text display page provided by an exemplary embodiment of the present application;
图11为本申请示例性实施例提供的另一种商品推荐文本生成方法的流程示意图；Figure 11 is a schematic flow chart of another product recommendation text generation method provided by an exemplary embodiment of the present application;
图12为本申请示例性实施例提供的一种商品推荐文本生成方法的整体技术链路示意图；Figure 12 is a schematic diagram of the overall technical links of a product recommendation text generation method provided by an exemplary embodiment of the present application;
图13为本申请示例性实施例提供的一种商品推荐文本生成装置的结构示意图；Figure 13 is a schematic structural diagram of a product recommendation text generation device provided by an exemplary embodiment of the present application;
图14为本申请示例性实施例提供的另一种商品推荐文本生成装置的结构示意图；Figure 14 is a schematic structural diagram of another product recommendation text generation device provided by an exemplary embodiment of the present application;
图15为本申请示例性实施例提供的另一种商品推荐文本生成装置的结构示意图；Figure 15 is a schematic structural diagram of another product recommendation text generation device provided by an exemplary embodiment of the present application;
图16为本申请示例性实施例提供的另一种商品推荐文本生成装置的结构示意图；Figure 16 is a schematic structural diagram of another product recommendation text generation device provided by an exemplary embodiment of the present application;
图17为本申请示例性实施例提供的一种电子设备的结构示意图。Figure 17 is a schematic structural diagram of an electronic device provided by an exemplary embodiment of the present application.
具体实施方式Detailed ways
为使本申请的目的、技术方案和优点更加清楚，下面将结合本申请具体实施例及相应的附图对本申请技术方案进行清楚、完整地描述。显然，所描述的实施例仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。本申请所涉及的用户信息(包括但不限于用户设备信息、用户个人信息等)和数据(包括但不限于用于分析的数据、存储的数据、展示的数据等)，均为经用户授权或者经过各方充分授权的信息和数据，并且相关数据的收集、使用和处理需要遵守相关国家和地区的相关法律法规和标准，并提供有相应的操作入口，供用户选择授权或者拒绝。In order to make the purpose, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with specific embodiments of the present application and corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application. The user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all authorized by the user or Information and data that have been fully authorized by all parties, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation portals for users to choose to authorize or refuse.
在线上购物过程中，由于商品种类和数量较多，为了提高消费者的点击量、浏览量以及购买量等，电商类APP通常会在显示商品的同时显示商品的推荐理由，例如商品推荐文本(或者文案)等。在相关技术中，商品推荐文本通常是由运营人员人工配置生成，但由于商品数量众多，运营人员需要逐个对各个商品人工配置商品推荐文本，同时由于商品数据是动态变化的，运营人员还需要人工对商品推荐文本进行动态调整。这种商品推荐文本的配置方式人力资源耗费较多，时间耗费较长，同时运营人员基于自身经验人工配置的商品推荐文本准确性也不高，无法契合消费者的实际需求，对消费者的吸引力不强。During the online shopping process, due to the large number of product types and quantities, in order to increase consumer clicks, views, and purchases, e-commerce APPs usually display the product recommendation reasons, such as product recommendation text, while displaying the product. (or copywriting) etc. In related technologies, product recommendation texts are usually manually configured and generated by operators. However, due to the large number of products, operators need to manually configure product recommendation texts for each product one by one. At the same time, because product data changes dynamically, operators also need to manually configure product recommendation texts. Dynamically adjust product recommendation text. This method of configuring product recommendation text consumes a lot of human resources and takes a long time. At the same time, the product recommendation text manually configured by operators based on their own experience is not very accurate and cannot meet the actual needs of consumers. It is not attractive to consumers. Not strong.
图1为本申请示例性实施例提供的一种应用场景的示意图。如图1所示，包括电子设备101以及运营人员102。电子设备101具体可以为手机、电脑、可穿戴设备等，本申请实施例对于电子设备的具体种类不作限定。Figure 1 is a schematic diagram of an application scenario provided by an exemplary embodiment of the present application. As shown in Figure 1, it includes electronic equipment 101 and operating personnel 102. The electronic device 101 may specifically be a mobile phone, a computer, a wearable device, etc. The embodiment of the present application does not limit the specific type of the electronic device.
如图1所示出的，在相关技术中，针对众多商品，运营人员102需要人工逐个生成商品推荐文本，逐个进行商品文案的编辑和输入操作，这种商品推荐文本的生成方式人力资源耗费较大，耗费了运营人员102大量的时间。而在本申请实施例中，电子设备101可以基于运营人员102的简单配置实现商品推荐文本的自动生成，无需运营人员102针对众多商品逐个进行人工输入，能够节省人力资源和时间成本。As shown in Figure 1, in the related technology, for many products, the operator 102 needs to manually generate product recommendation text one by one, and edit and input the product copy one by one. This method of generating product recommendation text consumes more human resources. It is large and consumes a lot of time of the operation staff 102. In the embodiment of the present application, the electronic device 101 can automatically generate product recommendation text based on simple configuration by the operator 102, without the need for the operator 102 to manually input numerous products one by one, which can save human resources and time costs.
本申请实施例中，服务端获取第一关联数据、第二关联数据以及目标格式文本；第一关联数据包括商品标识以及商品标签；第二关联数据包括用户标识以及人群标签；目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；根据第二关联数据，创建人群标签对应的人群标识并确定人群标识包括的用户标识；在目标格式文本中包括行为数据的情况下，获取商品标识在时间周期内对应的历史数据，并根据历史数据、人群标识以及行为数据，确定人群标识对应的目标行为数据；基于目标行为数据，按照目标格式文本生成商品标识对应的商品推荐文本。本申请实施例中，服务端可以根据运营人员配置的目标格式文本来自动统计目标行为数据并自动生成、更新商品推荐文本，无需运营人员进行文案的人工生成，能够降低人力资源的耗费，节省时间成本，同时服务端可以根据运营人员上传的第二关联数据来自动创建人群标识，并且将人群标识与多个用户标识相关联，这样能够快速确定消费者的所属的人群标识，结合多样化的商品标签，可以提高文案推荐文本的准确性，更加契合消费者的实际需求，能够增强对消费者的吸引力。In the embodiment of this application, the server obtains the first associated data, the second associated data and the target format text; the first associated data includes product identification and product tags; the second associated data includes user identification and group tags; the target format text includes pre- Suppose there are at least two types of description text, product tag, time period and behavioral data; according to the second associated data, create a crowd identifier corresponding to the crowd tag and determine the user identifier included in the crowd identifier; when the target format text includes behavioral data , obtain the historical data corresponding to the product identification within the time period, and determine the target behavior data corresponding to the crowd identification based on historical data, crowd identification, and behavioral data; based on the target behavior data, generate product recommendation text corresponding to the product identification according to the target format text . In the embodiment of this application, the server can automatically count target behavior data and automatically generate and update product recommendation text based on the target format text configured by the operator. There is no need for the operator to manually generate copywriting, which can reduce the consumption of human resources and save time. cost, and at the same time, the server can automatically create a crowd ID based on the second associated data uploaded by the operator, and associate the crowd ID with multiple user IDs, so that the consumer's crowd ID can be quickly determined and combined with diversified products Tags can improve the accuracy of copywriting recommendation text, better meet the actual needs of consumers, and enhance the appeal to consumers.
以下通过具体实施例对本申请所示的技术方案进行详细说明。需要说明的是，以下几个实施例可以单独存在，也可以相互结合，对于相同或相似的内容，在不同的实施例中不再重复说明。The technical solutions shown in this application will be described in detail below through specific examples. It should be noted that the following embodiments can exist alone or can be combined with each other. The same or similar content will not be repeatedly described in different embodiments.
图2为本申请示例性实施例提供的一种商品推荐文本生成方法的流程示意图。请参见图2，该商品推荐文本生成方法可以包括：FIG. 2 is a schematic flowchart of a method for generating product recommendation text provided by an exemplary embodiment of the present application. See Figure 2. The product recommendation text generation method may include:
S201、获取第一关联数据、第二关联数据以及目标格式文本；第一关联数据包括商品标识以及商品标签；第二关联数据包括用户标识以及人群标签；目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种。S201. Obtain the first associated data, the second associated data and the target format text; the first associated data includes product identification and product tags; the second associated data includes user identification and group tags; the target format text includes preset description text and product tags. , time period and behavioral data at least two of them.
本申请实施例的执行主体可以为服务端，例如服务器等，也可以为设置在服务端中的商品推荐文本生成装置。商品推荐文本生成装置可以通过软件实现，也可以通过软件和硬件的结合实现。为了便于理解，在下文中，以执行主体为服务端为例进行说明。The execution subject of the embodiment of the present application may be a server, such as a server, or may be a product recommendation text generating device provided in the server. The product recommendation text generating device can be implemented by software or a combination of software and hardware. For ease of understanding, in the following description, the execution subject is the server as an example.
需要说明的是，在本申请实施例中，服务端可以用于进行商品推荐文本的生成和确定，具体可以是指服务器等电子设备；客户端既可以是指运营人员进行商品推荐文本配置时所使用的客户端，也可以是指消费者访问电商类APP中特定页面时所使用的客户端，具体可以是指手机、电脑等电子设备，本申请实施例对于服务端以及客户端的具体种类不作限定。第一用户可以是指运营人员、运维人员、数据统计人员或者开发人员等。第二用户可以是指消费者等，当然，在一些场景中，第一用户和第二用户也可以为同一用户，本申请实施例对此不作限定。It should be noted that in the embodiment of the present application, the server can be used to generate and determine the product recommendation text, and specifically can refer to electronic equipment such as a server; the client can either refer to the operator when configuring the product recommendation text. The client used may also refer to the client used by consumers when accessing specific pages in e-commerce APPs. Specifically, it may refer to electronic devices such as mobile phones and computers. The embodiments of this application do not specify the specific types of servers and clients. limited. The first user may refer to operations personnel, operation and maintenance personnel, data statisticians or developers, etc. The second user may refer to a consumer or the like. Of course, in some scenarios, the first user and the second user may also be the same user, which is not limited in the embodiments of this application.
本申请实施例中，第一关联数据可以是指商品的具体数据，其中可以包括具备关联关系的用户标识-商品标识-商品标签等，该关联关系可以为一对多的关系，例如一个用户标识可以对应多个商品标识，一个商品标识可以对应多个用户标识，一个商品标识可以对应多个商品标签等，当然也可以包括其他关联关系，本申请实施例对于关联关系的具体形式不作限定。具体的，用户标识可以是指消费者的用户标识(User Identification，UID)等。商品标识可以是指商品的唯一标识等，具体可以是指商品编号等。第一关联数据中商品标识对应的商品可以是指第一用户预先选定的商品，例如可以是指从待推广的商品池中选取的商品等。商品标签可以是指商品对应的各种标签名称，具体可以是指地理位置标签、季节时令标签、身份标签或者类别标签等，当然，商品标签还可以包括其他类型的标签，本申请实施例对此不作限定。In the embodiment of this application, the first associated data may refer to the specific data of the product, which may include an associated relationship of user identification - product identification - product tag, etc. The association relationship may be a one-to-many relationship, such as a user identification It can correspond to multiple product identifiers, one product identifier can correspond to multiple user identifiers, one product identifier can correspond to multiple product tags, etc., and of course it can also include other associated relationships. The embodiment of this application does not limit the specific form of the associated relationship. Specifically, the user identification may refer to the consumer's user identification (User Identification, UID), etc. The product identification may refer to the unique identification of the product, etc. Specifically, it may refer to the product number, etc. The product corresponding to the product identifier in the first associated data may refer to a product pre-selected by the first user, for example, may refer to a product selected from a product pool to be promoted, etc. Product tags may refer to various tag names corresponding to the product, specifically, they may refer to geographical location tags, seasonal tags, identity tags or category tags, etc. Of course, product tags may also include other types of tags. In this regard, the embodiment of the present application Not limited.
第二关联数据可以是指人群的具体数据，其中可以包括具备关联关系的用户标识-人群标签等。人群标签可以是指具备相同特征的一类人群对应的标签名称，具体可以是指地理位置标签、身份标签等，例如可以是“广东用户”、“浙江用户”、“学生党”等，当然人群标签也可以包括其他类型，本申请实施例对此不作限定。The second associated data may refer to specific data of a group of people, which may include associated user identifiers-group tags, etc. Crowd tags can refer to tag names corresponding to a group of people with the same characteristics. Specifically, they can refer to geographical location tags, identity tags, etc., for example, they can be "Guangdong users", "Zhejiang users", "Student Party", etc. Of course, crowds Tags may also include other types, which are not limited in the embodiments of this application.
目标格式文本可以是指第一用户预先配置的文案模版，可以是预设描述文本、商品标签、时间周期以及行为数据其中至少两种文本元素的组合。其中，预设描述文本可以是指描述性的自定义文本词语，例如“偏爱”、“喜爱”、“喜欢”等。商品标签可以是指第一用户选择展示的标签名称。时间周期可以是指预先设置的数据统计周期，例如可以是7天、30天、60天等。行为数据可以包括具体数值与行为类型，用于表征某个消费行为的统计数量，例如可以是指“600+购买”、“3000+浏览”等。The target format text may refer to a copy template preconfigured by the first user, and may be a combination of at least two text elements among preset description text, product tags, time period and behavioral data. The preset description text may refer to descriptive custom text words, such as "preference", "favorite", "like", etc. The product tag may refer to the tag name selected by the first user to display. The time period may refer to a preset data statistics period, for example, it may be 7 days, 30 days, 60 days, etc. Behavior data can include specific values and behavior types, used to represent the statistical quantity of a certain consumption behavior, for example, it can refer to "600+ purchases", "3000+ views", etc.
具体的，目标格式文本可以是上述四种文本元素的组合，也可以是其中任意两个或者三个文本元素的组合，当然还可以是其中任意一个文本元素。示例性地，当目标格式文本为商品标签、预设描述文本的组合时，服务端最终生成的商品推荐文本可以为“浙江用户喜欢”，其中“浙江用户”为商品标签，“喜欢”为预设描述文本；当目标格式文本为商品标签、时间周期的组合时，最终生成的商品推荐文本可以为“连衣裙新品”，其中“近7内”为时间周期，可以显示也可以不显示，“连衣裙新品”为预设描述文本；当目标格式文本为商品标签、行为数据的组合时，最终生成的商品推荐文本可以为“5000+浙江用户购买”，其中“浙江用户”为商品标签，“5000+购买”为行为数据，此时也可以同时显示时间周期等。当然，目标格式文本的具体组合方式也可以为其他形式，还可以包括其他更多元素，具体可以基于实际需求进行灵活设置，本申请实施例对此不作限定。Specifically, the target format text can be a combination of the above four text elements, or a combination of any two or three text elements, and of course it can be any one of the text elements. For example, when the target format text is a combination of product tags and preset description text, the product recommendation text finally generated by the server can be "Zhejiang users like", where "Zhejiang users" is the product tag and "like" is the preset description text. Assume description text; when the target format text is a combination of product tags and time periods, the final product recommendation text can be "Dress New Product", where "in the past 7 years" is the time period, which can be displayed or not, and "Dress "New product" is the preset description text; when the target format text is a combination of product tags and behavioral data, the final product recommendation text can be "5000+ Zhejiang users purchased", where "Zhejiang users" is the product tag, and "5000+ Purchase" is behavioral data, and the time period, etc. can also be displayed at this time. Of course, the specific combination method of the target format text can also be in other forms, and can also include other more elements. The specific combination can be flexibly set based on actual needs, which is not limited in the embodiments of this application.
在本步骤中，第一用户可以在客户端进行商品推荐文本的统一配置，服务端后续可以基于第一用户的配置内容自动生成商品推荐文本。具体的，第一用户可以选择第一关联数据以及第二关联数据，服务端可以基于第一关联数据读取商品标识和商品标签，并且可以基于第二关联数据来自动批量创建人群标识，简化第一用户的操作。并且，第一用户可以预先配置商品推荐文本对应的文案模版，即目标格式文本，服务端后续可以基于该目标格式文本自动生成商品推荐文本，这样无需第一用户针对每个商品来人工逐个输入推荐文案，能够减少人力成本和时间成本。In this step, the first user can uniformly configure the product recommendation text on the client, and the server can subsequently automatically generate the product recommendation text based on the first user's configuration content. Specifically, the first user can select the first associated data and the second associated data, the server can read the product identifiers and product tags based on the first associated data, and can automatically create crowd identifiers in batches based on the second associated data, simplifying the A user's action. Moreover, the first user can pre-configure the copy template corresponding to the product recommendation text, that is, the target format text. The server can subsequently automatically generate the product recommendation text based on the target format text. This eliminates the need for the first user to manually input recommendations for each product one by one. Copywriting can reduce labor costs and time costs.
S202、根据第二关联数据，创建人群标签对应的人群标识并确定人群标识包括的用户标识。S202. According to the second associated data, create a crowd identifier corresponding to the crowd tag and determine the user identifier included in the crowd identifier.
本申请实施例中，人群标识可以是指人群标签对应的唯一标识符，具体可以是指人群编号等。具体的，服务端在获取到第二关联数据之后，可以提取其中的人群标签，并自动、批量创建该人群标签对应的人群标识，同时可以确定各个人群标识所包括的多个用户标识，这样，相较于相关技术中第一用户需要逐个手动创建人群标识的方式，本申请实施例中服务端可以基于第二关联数据来实现逻辑人群的自动批量创建，能够简化第一用户的操作，节省第一用户的时间。In the embodiment of this application, the crowd identifier may refer to the unique identifier corresponding to the crowd tag, and specifically may refer to the crowd number, etc. Specifically, after obtaining the second associated data, the server can extract the crowd tags in it, and automatically and batch create crowd identifiers corresponding to the crowd tags. At the same time, it can determine multiple user identifiers included in each crowd identifier. In this way, Compared with the method in the related art in which the first user needs to manually create crowd identifiers one by one, in the embodiment of the present application, the server can realize the automatic batch creation of logical crowds based on the second associated data, which can simplify the first user's operation and save the second time. one user's time.
S203、在目标格式文本中包括行为数据的情况下，获取商品标识在时间周期内对应的历史数据，并根据历史数据、人群标识以及行为数据，确定人群标识对应的目标行为数据。S203. When the target format text includes behavioral data, obtain the historical data corresponding to the product identification within the time period, and determine the target behavior data corresponding to the crowd identification based on the historical data, crowd identification, and behavioral data.
本申请实施例中，历史数据可以是指消费者即第二用户对应的消费数据，该历史数据可以是指大量消费者对应的消费数据。目标行为数据可以是指人群标识对应的行为数据统计值。In the embodiment of this application, the historical data may refer to the consumption data corresponding to the consumer, that is, the second user, and the historical data may refer to the consumption data corresponding to a large number of consumers. The target behavior data may refer to the statistical value of behavior data corresponding to the crowd identification.
本步骤中，在目标格式文本包括行为数据的情况下，服务端需要统计各个人群标识对应的目标行为数据。服务端可以首先获取商品标识在时间周期内的历史数据，例如可以是获取30天内的历史数据，同时在历史数据中，基于行为数据的数据类型，统计人群标识对应的具体数值，得到目标行为数据。这样，本申请实施例中，服务端可以在历史数据中实现对目标行为数据的统计更新，能够提高商品推荐文本的准确性。In this step, when the target format text includes behavioral data, the server needs to count the target behavioral data corresponding to each group identifier. The server can first obtain the historical data of the product identification within the time period, for example, it can obtain the historical data within 30 days. At the same time, in the historical data, based on the data type of behavioral data, the specific values corresponding to the crowd identification are counted to obtain the target behavior data. . In this way, in this embodiment of the present application, the server can implement statistical updates of target behavior data in historical data, which can improve the accuracy of product recommendation text.
S204、基于目标行为数据，按照目标格式文本生成商品标识对应的商品推荐文本。S204. Based on the target behavior data, generate product recommendation text corresponding to the product identification according to the target format text.
本申请实施例中，商品推荐文本可以是指商品标识对应的推荐文案。本步骤中，服务端在确定出目标行为数据之后，可以将目标行为数据添加至目标格式文本中，最终得到商品标识对应的商品推荐文本。需要说明的是，当目标格式文本中不包括行为数据时，服务端可以无需再统计行为数据，直接基于目标格式文本中的元素来生成商品标识针对不同人群标识的商品推荐文本，本申请实施例对此不作限定。In the embodiment of this application, the product recommendation text may refer to the recommendation copy corresponding to the product identification. In this step, after determining the target behavior data, the server can add the target behavior data to the target format text, and finally obtain the product recommendation text corresponding to the product identification. It should be noted that when the target format text does not include behavioral data, the server can directly generate product identification and product recommendation text for different groups of people based on the elements in the target format text without counting the behavioral data. Embodiments of this application There is no limit to this.
服务端后续可以对商品标识-人群标识-商品推荐文本进行关联存储。在实际投放过程中，当第二用户对应的推荐商品命中第一关联数据中的商品标识时，服务端可以通过查询的方式确定出第二用户对应的商品推荐文本，进而在商品展示页面同步显示该商品推荐文本，能够实现商品推荐文本的精准投放，增强对第二用户的吸引力。The server can subsequently associate and store the product identification-crowd identification-product recommendation text. During the actual delivery process, when the recommended product corresponding to the second user hits the product identifier in the first associated data, the server can determine the product recommendation text corresponding to the second user through query, and then display it simultaneously on the product display page This product recommendation text can achieve accurate placement of product recommendation text and enhance the appeal to second users.
在本申请实施例中，服务端获取第一关联数据、第二关联数据以及目标格式文本；第一关联数据包括商品标识以及商品标签；第二关联数据包括用户标识以及人群标签；目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；根据第二关联数据，创建人群标签对应的人群标识并确定人群标识包括的用户标识；在目标格式文本中包括行为数据的情况下，获取商品标识在时间周期内对应的历史数据，并根据历史数据、人群标识以及行为数据，确定人群标识对应的目标行为数据；基于目标行为数据，按照目标格式文本生成商品标识对应的商品推荐文本。本申请实施例中，服务端可以根据运营人员配置的目标格式文本来自动统计目标行为数据并自动生成、更新商品推荐文本，无需运营人员进行文案的人工生成，能够降低人力资源的耗费，节省时间成本，同时服务端可以根据运营人员上传的第二关联数据来自动创建人群标识，并且将人群标识与多个用户标识相关联，这样能够快速确定消费者的所属的人群标识，结合多样化的商品标签，可以提高文案推荐文本的准确性，更加契合消费者的实际需求，能够增强对消费者的吸引力。In the embodiment of this application, the server obtains the first associated data, the second associated data and the target format text; the first associated data includes product identification and product tags; the second associated data includes user identification and group tags; the target format text includes Preset at least two of description text, product tag, time period and behavioral data; create a crowd identifier corresponding to the crowd tag according to the second associated data and determine the user ID included in the crowd identifier; include behavior data in the target format text Next, obtain the historical data corresponding to the product identification within the time period, and determine the target behavior data corresponding to the crowd identification based on historical data, crowd identification, and behavioral data; based on the target behavior data, generate product recommendations corresponding to the product identification according to the target format text text. In the embodiment of this application, the server can automatically count target behavior data and automatically generate and update product recommendation text based on the target format text configured by the operator. There is no need for the operator to manually generate copywriting, which can reduce the consumption of human resources and save time. cost, and at the same time, the server can automatically create a crowd ID based on the second associated data uploaded by the operator, and associate the crowd ID with multiple user IDs, so that the consumer's crowd ID can be quickly determined and combined with diversified products Tags can improve the accuracy of copywriting recommendation text, better meet the actual needs of consumers, and enhance the appeal to consumers.
在上述实施例的基础上，图3为本申请示例性实施例提供的另一种商品推荐文本生成方法的流程示意图。请参见图3，该商品推荐文本生成方法可以包括：Based on the above embodiments, FIG. 3 is a schematic flowchart of another method for generating product recommendation text provided by an exemplary embodiment of the present application. Please refer to Figure 3. The product recommendation text generation method may include:
S301、获取第一关联数据、第二关联数据以及目标格式文本；第一关联数据包括商品标识以及商品标签；第二关联数据包括用户标识以及人群标签；目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种。S301. Obtain the first associated data, the second associated data and the target format text; the first associated data includes product identification and product tags; the second associated data includes user identification and group tags; the target format text includes preset description text and product tags. , time period and behavioral data at least two of them.
在一种可能的实施方式中，商品标签包括地理位置标签、季节时令标签、身份标签、类别标签其中至少一种。In a possible implementation, the product tag includes at least one of a geographical location tag, a seasonal tag, an identity tag, and a category tag.
本申请实施例中，第一关联数据中的商品标签可以是第一用户基于实际需求灵活配置的，可以包括地理位置标签、季节时令标签、身份标签以及类别标签其中至少一种。其中，地理位置标签可以用于表征商品的地理区域特征，例如具体可以是“北京特产”、“杭州特产”等。季节时令标签可以用于表征商品的节令属性特征，例如具体可以是“春节礼品”、“中秋礼品”等。身份标签可以用于表征商品对应的适用人群等，例如可以是“上班族”等。类别标签可以用于表征商品的具体类别，例如可以是“萌宠用品”、“居家清扫”等。当然，除了上述标签之外，商品标签还可以包括其他类型或者其他内容的标签，例如“新品”、“折扣”等商家活动标签等，具体可以基于实际需求进行灵活设置，本申请实施例对此不作限定。In this embodiment of the present application, the product tags in the first associated data may be flexibly configured by the first user based on actual needs, and may include at least one of a geographical location tag, a seasonal tag, an identity tag, and a category tag. Among them, the geographical location tag can be used to characterize the geographical regional characteristics of the product, for example, it can be "Beijing specialties", "Hangzhou specialties", etc. Seasonal tags can be used to characterize seasonal attributes of goods, for example, they can be "Spring Festival gifts", "Mid-Autumn Festival gifts", etc. The identity tag can be used to characterize the applicable group of products, such as "office workers". Category tags can be used to represent specific categories of products, such as "cute pet supplies", "home cleaning", etc. Of course, in addition to the above tags, product tags can also include tags of other types or other content, such as "new product", "discount" and other merchant activity tags, etc., which can be flexibly set based on actual needs. This is the case in the embodiment of this application. Not limited.
本申请实施例中，服务端在生成商品推荐文本时，可以使用多样化的商品标签，能够实现商品推荐的个性化扩展，提高了商品推荐文本的丰富性，能够更加契合消费者的实际需求和个性偏好，进一步提高了对消费者的吸引力。In the embodiment of this application, the server can use a variety of product tags when generating product recommendation text, which can realize personalized expansion of product recommendation, improve the richness of product recommendation text, and can better meet the actual needs and needs of consumers. Personal preferences further enhance the appeal to consumers.
S302、读取第二关联数据中的多个人群标签，分别创建各个人群标签对应的人群标识；根据第二关联数据中用户标识与人群标签的关联关系，确定各个人群标识包括的用户标识，并且存储人群标识以及人群标识包括的用户标识。S302. Read multiple crowd tags in the second associated data, and create crowd identifiers corresponding to each crowd tag respectively; determine the user identifiers included in each crowd identifier according to the association between user identifiers and crowd tags in the second associated data, and Store crowd identification and user IDs included in the crowd identification.
本申请实施例中，第二关联数据可以是指人群标签表，其中包括有用户标识以及人群标签。服务端在获取到第二关联数据之后，可以读取其中的人群标签，并创建各个人群标签对应的人群标识，然后服务端可以基于人群标签表中的关联关系，确定出各个人群标识包括的用户标识。这样，服务端基于第二关联数据，能够实现人群标识的自动、批量创建，无需第一用户手动逐个点击创建，能够简化第一用户的操作，节省第一用户的时间，同时也能够降低手动创建人群标识时的出错概率，提高人群标识创建的准确度。In this embodiment of the present application, the second associated data may refer to a crowd tag table, which includes user identifiers and crowd tags. After obtaining the second associated data, the server can read the crowd tags in it and create crowd identifiers corresponding to each crowd tag. Then the server can determine the users included in each crowd tag based on the association relationships in the crowd tag table. logo. In this way, based on the second associated data, the server can realize automatic and batch creation of crowd identifiers without the need for the first user to manually click to create one by one, which can simplify the first user's operation, save the first user's time, and also reduce the number of manual creations. The error probability during crowd identification improves the accuracy of crowd identification creation.
具体的，服务端在批量创建人群标识时，可以首先将第二关联数据(人群标签表)写入缓存中，然后可以查询人群标签表中的标签字段，读取出多个人群标签；服务端针对各个人群标签可以创建逻辑人群，确定各个人群标签对应的人群标识，示例性地，假设人群标签表中包括有30个人群标签，服务端可以针对每个人群标签确定对应的人群标识。之后，服务端可以基于人群标签表中的关联关系，确定出各个人群标识对应的用户标识，并且可以将具备映射关系的人群标识与用户标识存储至预设数据表中，实现人群标识与用户标识的稳定存储。在此基础上，后续服务端在获取到第二用户的用户标识之后，可以基于该映射关系快速确定出第二用户对应的人群标识，能够实现对消费者人群特征的快速确定，使得商品推荐文本能够更加契合消费者的个性偏好，可以提高商品推荐文本的准确性。Specifically, when the server creates crowd identifiers in batches, it can first write the second associated data (crowd tag table) into the cache, and then query the tag field in the crowd tag table and read out multiple crowd tags; the server A logical crowd can be created for each crowd tag, and the crowd identifier corresponding to each crowd tag can be determined. For example, assuming that the crowd tag table includes 30 crowd tags, the server can determine the corresponding crowd identifier for each crowd tag. After that, the server can determine the user ID corresponding to each crowd ID based on the association relationship in the crowd tag table, and can store the crowd ID and user ID with mapping relationship into the preset data table to realize the crowd ID and user ID. stable storage. On this basis, after the subsequent server obtains the user identification of the second user, it can quickly determine the group identification corresponding to the second user based on the mapping relationship, which can quickly determine the characteristics of the consumer group and make product recommendation text It can be more in line with consumers' personal preferences and improve the accuracy of product recommendation text.
S303、周期性检测第一关联数据与第二关联数据的更新状态；在第一关联数据与第二关联数据更新完成后，获取商品标识在时间周期内的历史数据；历史数据包括商品标识对应的历史行为数据。S303. Periodically detect the update status of the first associated data and the second associated data; after the update of the first associated data and the second associated data is completed, obtain the historical data of the product identification within the time period; the historical data includes the product identification corresponding to Historical behavioral data.
本申请实施例中，更新状态可以是指第一关联数据与第二关联数据的实时状态。第一关联数据可以是指用户标识-商品标识-商品标签表，第二关联数据可以是指人群标签表，第一关联数据和第二关联数据可以是动态更新的，具体可以是12小时、24小时或者48小时更新一次等，具体更新方式可以是由服务端自动更新，也可以是通过第一用户来手动更新，本申请实施例对此不作限定。历史行为数据可以是指消费者对应的历史消费行为数据，具体可以是指浏览行为、收藏行为、购买行为、加购(加入购物车)行为等。In this embodiment of the present application, the update status may refer to the real-time status of the first associated data and the second associated data. The first associated data may refer to the user identification-product identification-product tag table, and the second associated data may refer to the crowd tag table. The first associated data and the second associated data may be dynamically updated, specifically, they may be 12 hours, 24 hours, or 24 hours. It is updated once every hour or every 48 hours, etc. The specific update method may be automatically updated by the server, or it may be manually updated by the first user, which is not limited in the embodiments of the present application. Historical behavior data may refer to historical consumption behavior data corresponding to consumers, specifically, it may refer to browsing behavior, collection behavior, purchasing behavior, additional purchase (add to shopping cart) behavior, etc.
在本步骤中，服务端获取到第一关联数据、第二关联数据以及目标格式文本之后，可以创建有向无环图(Dag)的调度任务，以实现对商品推荐文本的实时更新。服务端基于Dag调度，可以将第一关联数据、第二关联数据的更新状态作为判断节点，周期性检测这两个数据表的更新状态，具体检测周期可以为1小时、6小时、12小时等。当这两个数据表完成更新之后，服务端可以基于Dag调度，执行商品推荐文本生成节点，此时服务端可以获取时间周期内商品标识对应的历史数据，后续可以基于历史数据以及目标格式文本来生成商品推荐文本。In this step, after the server obtains the first associated data, the second associated data and the target format text, it can create a scheduling task of a directed acyclic graph (Dag) to achieve real-time updating of the product recommendation text. Based on Dag scheduling, the server can use the update status of the first associated data and the second associated data as judgment nodes to periodically detect the update status of these two data tables. The specific detection period can be 1 hour, 6 hours, 12 hours, etc. . After the two data tables are updated, the server can execute the product recommendation text generation node based on Dag scheduling. At this time, the server can obtain the historical data corresponding to the product identification within the time period. Subsequently, it can be based on the historical data and the target format text. Generate product recommendation text.
在本申请实施例中，服务端周期性检测第一关联数据、第二关联数据的更新状态，当第一关联数据、第二关联数据对应的数据表完成更新之后，服务端可以获取历史数据并执行后续的商品推荐文本生成过程。这样，服务端基于第一关联数据、第二关联数据的更新状态来实现商品推荐文本的动态更新，能够实现商品推荐文本的自动更新，可以确保商品推荐文本的时效性，无需第一用户手动进行文案更新，能够减少第一用户的工作量。In the embodiment of this application, the server periodically detects the update status of the first associated data and the second associated data. After the data tables corresponding to the first associated data and the second associated data are updated, the server can obtain the historical data and Execute the subsequent product recommendation text generation process. In this way, the server realizes dynamic updating of the product recommendation text based on the update status of the first associated data and the second associated data, can realize automatic updating of the product recommendation text, and can ensure the timeliness of the product recommendation text without the need for the first user to manually perform Copywriting updates can reduce the workload of the first user.
S304、确定行为数据对应的目标行为类型；在历史数据中，确定人群标识对应的目标行为类型的数量，得到人群标识对应的目标行为数据。S304. Determine the target behavior type corresponding to the behavior data; in the historical data, determine the number of target behavior types corresponding to the crowd identifier, and obtain the target behavior data corresponding to the crowd identifier.
本申请实施例中，目标行为类型可以是指第一用户预配置的、需要统计的消费者行为类型，例如购买行为等。服务端在获取到历史数据之后，可以确定出目标格式文本中行为数据对应的目标行为类型，然后在该历史数据中，统计人群标识对应的目标行为类型的数量，例如具体可以统计该人群标识中有多少个第二用户有购买行为等，得到人群标识对应的目标行为数据。这样，服务端通过统计消费者的历史行为数据，结合人群标识来确定目标行为数据，后续基于目标行为数据来生成商品推荐文本，更能够契合消费者的实际需求，可以增强对消费者的吸引力。In this embodiment of the present application, the target behavior type may refer to a consumer behavior type that is preconfigured by the first user and needs to be counted, such as purchasing behavior. After obtaining the historical data, the server can determine the target behavior type corresponding to the behavior data in the target format text, and then count the number of target behavior types corresponding to the crowd identifier in the historical data. For example, it can specifically count the number of target behavior types corresponding to the crowd identifier. How many second users have purchasing behavior, etc., to obtain the target behavior data corresponding to the crowd identification. In this way, the server determines the target behavior data by counting the historical behavior data of consumers and combining it with crowd identification. It then generates product recommendation text based on the target behavior data, which can better meet the actual needs of consumers and enhance its appeal to consumers. .
S305、基于目标行为数据，按照目标格式文本生成商品标识对应的商品推荐文本。S305. Based on the target behavior data, generate product recommendation text corresponding to the product identification according to the target format text.
本申请实施例中，在目标格式文本中包括行为数据的情况下，服务端可以获取历史数据并统计得到目标行为数据，并进一步基于目标行为数据对目标格式文本进行字段填充，得到商品推荐文本。在目标格式文本不包括行为数据的情况下，服务端则无需对目标行为数据进行统计确定，可以直接基于目标格式文本中的文本元素，针对商品标识生成商品推荐文本。此外，第一用户可以选择不采用历史数据，而是采用自定义的其他历史数据数据表，此时第一用户需要同时配置查询语句(或者查询条件)，服务端可以基于其他历史数据表以及查询语句来确定目标行为数据，并基于目标行为数据来生成商品推荐文本。In the embodiment of this application, when the target format text includes behavioral data, the server can obtain historical data and collect statistics to obtain the target behavior data, and further fill in the fields of the target format text based on the target behavior data to obtain product recommendation text. When the target format text does not include behavioral data, the server does not need to statistically determine the target behavior data. It can directly generate product recommendation text based on the product identification based on the text elements in the target format text. In addition, the first user can choose not to use historical data, but to use other customized historical data tables. At this time, the first user needs to configure query statements (or query conditions) at the same time, and the server can query based on other historical data tables and statements to determine the target behavior data, and generate product recommendation text based on the target behavior data.
本申请实施例中，服务端通过读取和计算历史数据的方式，结合目标格式文本，基于真实用户历史行为拼接出商品的商品推荐文本，无需第一用户进行人工操作，能够节约人工成本，提高了商品推荐文本的配置效率和准确率。In the embodiment of this application, the server reads and calculates historical data, combines the target format text, and splices the product recommendation text of the product based on the real user's historical behavior, without the need for the first user to perform manual operations, which can save labor costs and improve Improves the configuration efficiency and accuracy of product recommendation text.
S306、基于商品标识、人群标识以及商品推荐文本之间的关联关系，将商品标识、人群标识以及商品推荐文本存储至目标数据表中；接收客户端发送的数据导出请求，生成目标文件并向客户端发送目标文件。S306. Based on the association between the product identification, crowd identification and product recommendation text, store the product identification, crowd identification and product recommendation text into the target data table; receive the data export request sent by the client, generate the target file and send it to the customer Send the target file.
本申请实施例中，目标数据表可以是指预先设置的存储表，该目标数据表可以是表格的形式，当然也可以为其他形式，本申请实施例对此不作限定。数据导出请求可以是指客户端基于第一用户的数据导出操作向服务端发送的导出请求。目标文件可以是指包括人群标识、商品标识以及商品推荐文本的数据文件。In the embodiment of the present application, the target data table may refer to a preset storage table. The target data table may be in the form of a table, and of course may also be in other forms, which is not limited in the embodiment of the present application. The data export request may refer to an export request sent by the client to the server based on the first user's data export operation. The target file may refer to a data file including crowd identification, product identification, and product recommendation text.
在一些实际场景中，执行商品推荐文本生成任务的服务端与进行商品推荐的服务端可能不是同一个服务端，第一用户可能需要进行商品推荐文本的数据跨端传输等操作，相关技术中，第一用户通常是人工进行数据的录入，工作量较大，时间耗费较多。在本申请实施例中，服务端在按照目标格式文本的模版格式，基于第一关联数据、第二关联数据以及历史数据自动生成商品推荐文本之后，可以按照商品标识、人群标识以及商品推荐文本的关联关系，将商品标识、人群标识以及商品推荐文本存储至目标数据表中。在接收到客户端发送的数据导出请求的情况下，服务端可以基于目标数据表以及数据导出请求来导出目标文件，并向客户端发送该目标文件，实现数据的快速导出。这样，服务端基于客户端发送的数据导出请求生成目标文件并返回至客户端，能够实现目标文件的“一键导出”，后续可以直接基于目标文件进行数据的跨端上传，无需第一用户手动操作录入，能够简化第一用户的操作，能够提高效率和准确率。In some actual scenarios, the server that performs the product recommendation text generation task and the server that performs product recommendation may not be the same server. The first user may need to perform operations such as cross-end data transmission of product recommendation text. In related technologies, The first user usually enters data manually, which requires a large workload and consumes a lot of time. In this embodiment of the present application, after the server automatically generates the product recommendation text based on the first associated data, the second associated data and historical data according to the template format of the target format text, it can then generate the product recommendation text according to the product identification, crowd identification and product recommendation text. The association relationship stores product identification, crowd identification, and product recommendation text into the target data table. After receiving the data export request sent by the client, the server can export the target file based on the target data table and the data export request, and send the target file to the client to achieve rapid export of data. In this way, the server generates the target file based on the data export request sent by the client and returns it to the client, which can realize "one-click export" of the target file. Subsequent cross-end upload of data can be directly based on the target file without the need for the first user to manually Operation entry can simplify the first user's operations and improve efficiency and accuracy.
示例性地，图4为本申请示例性实施例提供的一种服务端商品推荐文本生成过程的技术链路图。如图4所示出的，服务端在获取到第一关联数据、第二关联数据以及目标格式文本等配置信息之后，可以创建用于生成商品推荐文本的Dag调度，该Dag调度依赖第一关联数据以及第二关联数据这两个数据表作为判断节点。服务端可以周期性检测第一关联数据以及第二关联数据的更新状态，并在这两个数据表更新完成之后，服务端基于Dag调度可以读取第一关联数据以及第二关联数据，并获取时间周期内的历史数据，基于第一用户所配置的行为数据或者自定义条件(其他历史数据以及查询语句等)从历史数据中确定出各个人群标识对应的目标行为数据，基于该目标行为数据以及目标格式文本拼接出需要执行的数据库语句(Structured Query Language，SQL)，进而得到商品推荐文本。Exemplarily, FIG. 4 is a technical link diagram of a server-side product recommendation text generation process provided by an exemplary embodiment of the present application. As shown in Figure 4, after obtaining the configuration information such as the first associated data, the second associated data, and the target format text, the server can create a Dag schedule for generating product recommendation text, and the Dag schedule relies on the first association The two data tables of data and second related data serve as judgment nodes. The server can periodically detect the update status of the first associated data and the second associated data, and after the two data tables are updated, the server can read the first associated data and the second associated data based on Dag scheduling, and obtain Historical data within the time period, based on the behavioral data configured by the first user or custom conditions (other historical data and query statements, etc.), the target behavioral data corresponding to each group identification is determined from the historical data, and based on the target behavioral data and The target format text is spliced into the database statement (Structured Query Language, SQL) that needs to be executed, and then the product recommendation text is obtained.
示例性地，若目标格式文本中的行为数据对应的目标行为类型为加购、浏览行为，时间周期为30天，服务端基于Dag调度可以计算第一关联数据中某个商品标识在各个人群标识中有过加购和浏览行为的总次数，具体数值可以用1000+、2000+等替换，得到目标行为数据，然后按照目标格式文本进行拼接，例如可以拼接商品标签以及预设描述文本等，即可得到商品标识对于各个人群标识的商品推荐文本。For example, if the target behavior type corresponding to the behavior data in the target format text is additional purchase and browsing behavior, and the time period is 30 days, the server can calculate based on Dag scheduling that a certain product identifier in the first associated data is in each group of people. The total number of times there have been additional purchases and browsing behaviors. The specific values can be replaced with 1000+, 2000+, etc. to obtain the target behavior data, and then spliced according to the target format text. For example, product labels and preset description texts can be spliced, that is, The product recommendation text of the product identification for each group of people can be obtained.
服务端在生成商品推荐文本之后，可以将商品标识、人群标签以及商品推荐文本进行关联存储，存储至目标数据表中，并向客户端发送执行成功的通知信息，具体存储方式可以是按照商品标识进行分表存储等，本申请实施例对此不作限定；服务端后续可以基于客户端发送的数据导出请求，基于目标数据表导出目标文件，例如表格文件等，实现数据的快速导出，无需第一用户进行人工输入，能够进一步提高效率，降低用户的工作量。After the server generates the product recommendation text, it can associate and store the product identification, crowd label and product recommendation text, store it in the target data table, and send a notification message of successful execution to the client. The specific storage method can be based on the product identification. Perform storage in separate tables, etc. This application embodiment does not limit this; the server can subsequently export target files based on the target data table based on the data export request sent by the client, such as table files, etc., to achieve rapid export of data without first Manual input by users can further improve efficiency and reduce user workload.
在上述实施例的基础上，图5为本申请示例性实施例提供的另一种商品推荐文本生成方法的流程示意图。请参见图5，该商品推荐文本生成方法可以包括：Based on the above embodiments, FIG. 5 is a schematic flowchart of another method for generating product recommendation text provided by an exemplary embodiment of the present application. See Figure 5. The product recommendation text generation method may include:
S501、响应于第一用户的第一触控操作，确定第一关联数据以及第二关联数据；第一关联数据包括商品标识以及商品标签；第二关联数据包括用户标识以及人群标签。S501. In response to the first user's first touch operation, determine first associated data and second associated data; the first associated data includes product identification and product tags; the second associated data includes user identification and crowd tags.
本申请实施例的执行主体可以为客户端，例如手机、电脑等，也可以为设置在客户端中的商品推荐文本生成装置。商品推荐文本生成装置可以通过软件实现，也可以通过软件和硬件的结合实现。为了便于理解，在下文中，以执行主体为客户端为例进行说明。The execution subject of the embodiment of the present application can be a client, such as a mobile phone, a computer, etc., or it can be a product recommendation text generating device provided in the client. The product recommendation text generating device can be implemented by software or a combination of software and hardware. For ease of understanding, in the following description, the execution subject is the client as an example.
本申请实施例中，第一用户在客户端中对商品推荐文本进行人工配置时，可以首先通过第一触控操作来选择第一关联数据以及第二关联数据。其中，第一关联数据可以是指用户标识-商品标识-商品标签表，可以用于表征商品的具体信息；第二关联数据可以是指人群标签表，可以用于表征用户标识与人群标签的关联关系。第一触控操作可以是指第一用户的点击操作、滑动操作或者语音输入操作等，本申请实施例对于第一触控操作的具体类型不作限定。In the embodiment of the present application, when the first user manually configures the product recommendation text in the client, he can first select the first associated data and the second associated data through the first touch operation. Among them, the first associated data may refer to the user identification-product identification-product tag table, which can be used to represent the specific information of the product; the second associated data may refer to the crowd tag table, which can be used to characterize the association between the user ID and the crowd tag. relation. The first touch operation may refer to the first user's click operation, sliding operation, voice input operation, etc. The embodiment of the present application does not limit the specific type of the first touch operation.
S502、响应于第一用户的第二触控操作，确定目标格式文本；目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种。S502. In response to the first user's second touch operation, determine the target format text; the target format text includes at least two of preset description text, product tags, time period and behavior data.
本申请实施例中，第二触控操作可以是指第一用户的文案模版配置操作。第一用户可以基于实际需求，从预设描述文本、商品标签、时间周期以及行为数据其中选择至少两种来组成目标格式文本。In this embodiment of the present application, the second touch operation may refer to the first user's copy template configuration operation. Based on actual needs, the first user can select at least two from preset description text, product tags, time periods and behavioral data to form the target format text.
在一种可能的实施方式中，步骤S502具体可以包括以下两种方式的其中任意一种来实现：In a possible implementation, step S502 may be implemented in any of the following two ways:
(1)、响应于第二触控操作，基于第二触控操作对应的文本元素生成目标格式文本。(1) In response to the second touch operation, generate target format text based on the text element corresponding to the second touch operation.
本申请实施例中，第一用户在配置目标格式文本时，可以选择不统计用户行为数据，例如可以选择商品标签以及预设描述文本来组成目标格式文本。第一用户在配置目标格式文本时，也可以选择直接基于默认的历史数据表来统计用户行为数据，此时第一用户可以配置时间周期以及行为数据，确定用户行为数据的统计时段和统计类型。客户端响应于第一用户的第二触控操作，可以基于第一用户选择的文本元素来组成目标格式文本，具体组合方式可以基于实际需求进行灵活设置，本申请实施例对此不作限定。In this embodiment of the present application, when configuring the target format text, the first user may choose not to collect user behavior data. For example, the first user may select product tags and preset description text to form the target format text. When configuring the target format text, the first user can also choose to directly count user behavior data based on the default historical data table. At this time, the first user can configure the time period and behavior data to determine the statistical period and statistical type of the user behavior data. In response to the first user's second touch operation, the client can compose the target format text based on the text elements selected by the first user. The specific combination method can be flexibly set based on actual needs, which is not limited in the embodiments of the present application.
(2)响应于第二触控操作，基于第二触控操作对应的文本元素生成目标格式文本，并确定目标格式文本对应的其他历史数据以及查询条件。(2) In response to the second touch operation, generate target format text based on the text element corresponding to the second touch operation, and determine other historical data and query conditions corresponding to the target format text.
本申请实施例中，第一用户在选择统计用户行为数据时，也可以采用自定义的其他历史数据，此时第一用户需要同时配置查询条件(查询语句，例如where语句等)以确保用户行为数据的正确读取。In the embodiment of this application, when the first user chooses to collect user behavior data, he can also use other customized historical data. In this case, the first user needs to configure query conditions (query statements, such as where statements, etc.) at the same time to ensure user behavior Correct reading of data.
在本申请实施例中，客户端可以响应于第一用户的第二触控操作，确定目标格式文本包括的文本元素，其中可以包括基于历史数据统计得到的行为数据，也可以包括基于其他历史数据统计得到的行为数据，还可以直接不进行行为数据的统计，这样，能够确保目标格式文本的多样性，使得文案模版更加多样化，能够提高商品推荐文本的丰富性。In this embodiment of the present application, the client may respond to the first user's second touch operation to determine the text elements included in the target format text, which may include behavioral data obtained based on historical data statistics, or may include behavioral data based on other historical data. The behavioral data obtained by statistics can also be collected directly without statistics of behavioral data. This can ensure the diversity of the target format text, make the copy templates more diverse, and improve the richness of the product recommendation text.
示例性地，图6为本申请示例性实施例提供的一种商品推荐文本配置页面的示意图。如图6所示出的，在商品推荐文本配置页面中，客户端显示有多个可交互控件。第一用户可以基于实际需求，输入商品推荐文本配置策略的名称和描述信息，用以区分不同的文案配置策略。第一用户可以确定第一关联数据以及第二关联数据对应的数据表，具体的，第一用户可以直接输入数据表的名称，也可以与新的数据表建立关联关系。“关联人群表的字段”这一控件可以用于显示第二关联数据中的人群标签，第一用户可以触控“获取人群标签”来显示第二关联数据中的人群标签，这样能够方便第一用户对需要创建的人群标识进行直观查看。“圈人状态id”这一控件可以用于输入人群标签创建过程中，逻辑人群的具体创建方式，每个圈人状态id可以对应不同的人群创建方式和存储方式。Exemplarily, FIG. 6 is a schematic diagram of a product recommendation text configuration page provided by an exemplary embodiment of the present application. As shown in Figure 6, in the product recommendation text configuration page, the client displays multiple interactive controls. The first user can input the name and description information of the product recommendation text configuration strategy based on actual needs to distinguish different copywriting configuration strategies. The first user can determine the data table corresponding to the first associated data and the second associated data. Specifically, the first user can directly input the name of the data table, or can establish an association relationship with a new data table. The "Fields of Related Crowd Table" control can be used to display the crowd tags in the second linked data. The first user can touch "Get Crowd Tags" to display the crowd tags in the second linked data. This can facilitate the first user. Users can visually view the crowd identification that needs to be created. The "circle status id" control can be used to input the specific creation method of the logical crowd during the crowd tag creation process. Each circle status id can correspond to different crowd creation methods and storage methods.
此外，第一用户可以对行为数据的具体统计方式进行配置，在需要进行行为数据统计时，第一用户可以配置时间周期、行为数据的目标行为类型，此时第一客户端默认采用历史数据进行统计；在需要进行行为数据统计时，第一用户也可以进行自定义配置，即第一用户可以选择基于其他历史数据表进行行为统计，第一用户可以关联其他历史数据表，同时配置查询条件；当然，第一用户也可以选择不统计行为数据。In addition, the first user can configure the specific statistical method of behavioral data. When behavioral data statistics are needed, the first user can configure the time period and the target behavior type of the behavioral data. At this time, the first client uses historical data by default. Statistics; when behavioral data statistics are needed, the first user can also perform customized configuration, that is, the first user can choose to perform behavioral statistics based on other historical data tables, and the first user can associate other historical data tables and configure query conditions at the same time; Of course, the first user can also choose not to collect behavioral data.
“推荐理由设置”控件可以用于进行目标格式文本的具体配置，如图6中所示出的，目标格式文本可以包括行为数据(behavior)、商品标签(label)以及预设描述文本，行为数据等待填充的字段可以通过占位符来表示。第一用户在商品推荐文本页面中配置完成后，可以点击“保存”按钮，客户端可以保存第一用户的各种配置信息。当需要服务端执行配置信息生成商品推荐文本时，第一用户可以点击“提交运行”按钮，客户端可以向服务端发送该页面中的配置信息，服务端可以基于这些配置信息来生成商品推荐文本。此外，客户端还可以提供针对配置信息的审核功能，当第一用户点击“提交运行”按钮之后，客户端可以将该页面中的配置信息发送至审核节点，由审核人员审核通过之后再向服务端发送，这样可以确保商品推荐文本的合理性。The "Recommendation Reason Settings" control can be used to specifically configure the target format text, as shown in Figure 6. The target format text can include behavior data (behavior), product label (label) and preset description text, behavior data Fields waiting to be filled can be represented by placeholders. After the first user completes the configuration on the product recommendation text page, he can click the "Save" button, and the client can save various configuration information of the first user. When the server needs to execute configuration information to generate product recommendation text, the first user can click the "Submit to Run" button, and the client can send the configuration information in the page to the server, and the server can generate product recommendation text based on these configuration information. . In addition, the client can also provide an audit function for configuration information. When the first user clicks the "Submit to Run" button, the client can send the configuration information in the page to the audit node, which will be reviewed and approved by the auditor before being sent to the service. This way, the product recommendation text can be ensured to be reasonable.
需要说明的是，图6中商品推荐文本配置页面中的控件和配置方式仅仅为示例，在该页面中也可以包括其他可交互控件以实现其他配置方式，本申请实施例对此不作限定。It should be noted that the controls and configuration methods in the product recommendation text configuration page in Figure 6 are only examples. This page may also include other interactive controls to implement other configuration methods, which is not limited in this embodiment of the application.
S503、向服务端发送第一关联数据、第二关联数据以及目标格式文本，以使得服务端基于第二关联数据创建人群标签对应的人群标识并确定人群标识包括的用户标识，并且基于目标格式文本生成商品标识对应的商品推荐文本。S503. Send the first associated data, the second associated data and the target format text to the server, so that the server creates a crowd identifier corresponding to the crowd tag based on the second associated data and determines the user identifier included in the crowd identifier, and based on the target format text Generate product recommendation text corresponding to the product identification.
本申请实施例中，客户端在确定出第一用户配置的第一关联数据、第二关联数据以及目标格式文本之后，可以发送至服务端，服务端可以基于第二关联数据来批量创建逻辑人群，同时可以基于目标格式文本来生成商品推荐文本，实现商品推荐文本的自动创建，减少商品文案配置的人力成本和时间成本。In the embodiment of this application, after the client determines the first associated data, the second associated data and the target format text configured by the first user, it can send it to the server, and the server can create logical groups in batches based on the second associated data. , at the same time, product recommendation text can be generated based on the target format text, realizing the automatic creation of product recommendation text, and reducing the labor cost and time cost of product copywriting configuration.
在一种可能的实施方式中，该商品推荐文本生成方法还可以包括如下步骤：In a possible implementation, the product recommendation text generation method may also include the following steps:
响应于第一用户的数据导出操作，向服务端发送数据导出请求，以使得服务端基于数据导出请求生成目标文件；接收服务端返回的目标文件；目标文件中包括具备关联关系的商品标识、人群标识以及商品推荐文本。In response to the first user's data export operation, send a data export request to the server, so that the server generates a target file based on the data export request; receive the target file returned by the server; the target file includes associated product identifiers and groups of people Logo and product recommendation text.
本申请实施例中，数据导出操作可以是指第一用户对于数据导出交互控件的触控操作。客户端响应于第一用户的数据导出操作，可以生成数据导出请求并向服务端发送该数据导出请求，服务端在获取到数据导出请求之后，可以基于目标数据表以及数据导出请求生成目标文件，该目标文件中包括有具备关联关系的商品标识、人群标识以及商品推荐文本。之后，客户端可以获取服务端返回的目标文件，这样能够实现商品推荐文本数据的“一键导出”，无需第一用户进行手动录入，能够提高第一用户的操作效率和准确度。In this embodiment of the present application, the data export operation may refer to the first user's touch operation on the data export interactive control. In response to the first user's data export operation, the client can generate a data export request and send the data export request to the server. After obtaining the data export request, the server can generate a target file based on the target data table and the data export request. The target file includes associated product identifiers, crowd identifiers, and product recommendation text. After that, the client can obtain the target file returned by the server, which can realize "one-click export" of product recommendation text data without the need for manual entry by the first user, which can improve the first user's operational efficiency and accuracy.
示例性地，图7为本申请示例性实施例提供的一种商品推荐文本导出页面的示意图。如图7所示出的，在商品推荐文本导出页面中，客户端可以显示多个配置策略对应的具体信息。第一用户可以通过页面中的搜索控件来进行配置策略的搜索，具体可以按照“策略名称”、“创建时间”、“运行状态”等维度进行搜索，这样能够提高配置策略查找的效率，当然该搜索控件中也可以包括其他搜索维度，本申请实施例对此不作限定。客户端针对各个配置策略，显示有多个操作控件，其中“详情”控件可以用于显示配置策略的具体内容；“编辑”控件可以用于编辑配置信息；“记录”控件可以用于查看配置策略的更新记录；“导出”控件可以用于导出目标文件。第一用户需要导出目标文件时，可以触控“导出”控件，客户端响应于第一用户的数据导出操作，可以向服务端发送数据导出请求，并接收服务端返回的目标文件，实现商品推荐文本数据文件的一键导出，能够简化第一用户的操作，节省第一用户的时间。Exemplarily, FIG. 7 is a schematic diagram of a product recommendation text export page provided by an exemplary embodiment of the present application. As shown in Figure 7, in the product recommendation text export page, the client can display specific information corresponding to multiple configuration strategies. The first user can search for configuration policies through the search control on the page. Specifically, they can search according to dimensions such as "policy name", "creation time", "running status", etc. This can improve the efficiency of configuration policy search. Of course, this The search control may also include other search dimensions, which are not limited in the embodiments of this application. The client displays multiple operation controls for each configuration policy. The "Details" control can be used to display the specific content of the configuration policy; the "Edit" control can be used to edit configuration information; and the "Record" control can be used to view the configuration policy. The update record; the "Export" control can be used to export the target file. When the first user needs to export the target file, he can touch the "Export" control. In response to the first user's data export operation, the client can send a data export request to the server and receive the target file returned by the server to implement product recommendations. One-click export of text data files can simplify the first user's operation and save the first user's time.
示例性地，图8为本申请示例性实施例提供的一种商品推荐文本配置的具体链路示意图。如图8所示出的，客户端可以响应于第一用户的第一触控操作，确定第一关联数据以及第二关联数据，第一关联数据具体可以为用户标识-商品标识-商品标签表，第二关联数据具体可以为人群标签表等。之后，客户端可以响应于第一用户的第二触控操作，确定第一用户选择的用户行为表，具体可以是默认的历史数据，也可以是自定义的其他历史数据，客户端进一步可以确定出第一用户确定的时间周期和行为数据(行为类型字段等)；当第一用户配置不统计行为数据时，客户端则无需确定用户行为表，客户端之后可以确定出第一用户所设置的目标格式文本，具体可以通过占位符的形式表示待填充的字段。当第一用户配置完成并点击提交运行之后，客户端可以将第一关联数据、第二关联数据以及目标格式文本发送至服务端，服务端可以基于这些配置信息来生成商品推荐文本并进行存储。当第一用户需要导出目标文件时，客户端可以向服务端发送数据导出请求，由服务端生成目标文件并返回至客户端。这样，第一用户在客户端中通过简单配置即可实现商品推荐文本的自动生成和动态更新，能够减少人力资源耗费，节省时间成本，同时基于历史数据进行行为字段的统计确认，也能够提高商品推荐文本的准确性，能够提高对消费者的吸引力。Exemplarily, FIG. 8 is a specific link diagram of a product recommendation text configuration provided by an exemplary embodiment of the present application. As shown in Figure 8, the client can determine the first associated data and the second associated data in response to the first user's first touch operation. The first associated data can specifically be a user identification-product identification-product tag table. , the second associated data may specifically be a crowd tag table, etc. Afterwards, in response to the first user's second touch operation, the client can determine the user behavior table selected by the first user, which can be the default historical data or other customized historical data. The client can further determine The time period and behavior data (behavior type field, etc.) determined by the first user are output; when the first user configures not to count behavior data, the client does not need to determine the user behavior table. The client can later determine the behavior data set by the first user. Target format text, which can specifically represent the fields to be filled in the form of placeholders. After the first user completes the configuration and clicks Submit to run, the client can send the first associated data, the second associated data and the target format text to the server, and the server can generate product recommendation text based on these configuration information and store it. When the first user needs to export the target file, the client can send a data export request to the server, and the server generates the target file and returns it to the client. In this way, the first user can automatically generate and dynamically update product recommendation text through simple configuration in the client, which can reduce human resource consumption and save time and cost. At the same time, statistical confirmation of behavior fields based on historical data can also improve product quality. The accuracy of recommended text can improve the appeal to consumers.
在上述实施例的基础上，图9为本申请示例性实施例提供的另一种商品推荐文本生成方法的流程示意图。请参见图9，该商品推荐文本生成方法可以包括：Based on the above embodiments, FIG. 9 is a schematic flowchart of another method for generating product recommendation text provided by an exemplary embodiment of the present application. See Figure 9. The product recommendation text generation method may include:
S901、响应于第二用户的页面加载操作，确定第二用户的用户标识。S901. In response to the second user's page loading operation, determine the user ID of the second user.
本申请实施例的执行主体可以为客户端，例如手机、电脑等，也可以为设置在客户端中的商品推荐文本生成装置。商品推荐文本生成装置可以通过软件实现，也可以通过软件和硬件的结合实现。为了便于理解，在下文中，以执行主体为客户端为例进行说明。The execution subject of the embodiment of the present application can be a client, such as a mobile phone, a computer, etc., or it can be a product recommendation text generating device provided in the client. The product recommendation text generating device can be implemented by software or a combination of software and hardware. For ease of understanding, in the following description, the execution subject is the client as an example.
本申请实施例中，第二用户可以是指消费者。页面加载操作可以是指第二用户加载商品推荐页面的操作。在电商类APP中，客户端可以通过一些商品推荐页面来向用户推荐商品，具体可以是电商类APP的首页、搜索结果页面或者“猜你喜欢”等页面。客户端响应于第二用户的页面加载操作，可以确定出第二用户的用户标识，用户标识具体可以通过第二用户当前登录账号进行确定，也可以基于第二用户的设备标识进行确定，本申请实施例对此不作限定。In this embodiment of the present application, the second user may refer to a consumer. The page loading operation may refer to the operation of the second user loading the product recommendation page. In e-commerce APPs, clients can recommend products to users through some product recommendation pages, which can be the homepage of e-commerce APPs, search results pages, or "guess you like" pages. In response to the second user's page loading operation, the client can determine the user identification of the second user. The user identification can be determined based on the current login account of the second user, or can also be determined based on the device identification of the second user. This application The embodiment does not limit this.
S902、向服务端发送用户标识，以使得服务端根据用户标识确定第二用户对应的人群标识以及目标商品标识，并基于人群标识以及目标商品标识从商品推荐文本中确定第二用户对应的目标商品推荐文本。S902. Send the user identification to the server, so that the server determines the crowd identification and the target product identification corresponding to the second user based on the user identification, and determines the target product corresponding to the second user from the product recommendation text based on the crowd identification and the target product identification. Recommended text.
S903、展示目标商品标识对应的目标商品以及目标商品推荐文本。S903. Display the target product corresponding to the target product identification and the target product recommendation text.
本申请实施例中，客户端在确定出第二用户的用户标识之后，可以向服务端发送该用户标识。服务端基于该用户标识，可以确定出第二用户的用户标识所属的人群标识，并且基于商品推荐算法确定出用户标识对应的目标商品标识，具体的商品推荐算法可以基于实际需求进行灵活设置，本申请实施例对此不作限定。之后，服务端可以基于人群标识、目标商品标识，可以从预先存储的目标数据表中确定出第二用户对应的目标商品推荐文本。该目标数据表中存储有商品标识-人群标识-商品推荐文本的关联关系。在确定出目标商品标识以及目标商品推荐文本之后，服务端可以将目标商品标识以及目标商品推荐文本返回至客户端，由客户端在商品推荐页面中进行显示。这样，服务端可以快速确定出第二用户的人群标签特征，并且确定出第二用户对应的目标商品标识和目标商品推荐文本，能够契合第二用户的个性偏好，提高商品推荐文本的准确度。In this embodiment of the present application, after determining the user identification of the second user, the client may send the user identification to the server. Based on the user identification, the server can determine the group identification to which the second user's user identification belongs, and determine the target product identification corresponding to the user identification based on the product recommendation algorithm. The specific product recommendation algorithm can be flexibly set based on actual needs. The application examples do not limit this. Afterwards, the server can determine the target product recommendation text corresponding to the second user from the pre-stored target data table based on the crowd identification and the target product identification. The target data table stores the association relationship between product identification-crowd identification-product recommendation text. After determining the target product identification and target product recommendation text, the server may return the target product identification and target product recommendation text to the client, and the client may display the target product identification and target product recommendation text on the product recommendation page. In this way, the server can quickly determine the crowd tag characteristics of the second user, and determine the target product identification and target product recommendation text corresponding to the second user, which can fit the second user's personal preferences and improve the accuracy of the product recommendation text.
需要说明的，服务端中预先存储的人群标识、商品推荐文本的确定过程可以参照前述实施例，本申请实施例在此不再赘述。It should be noted that the process of determining the crowd identification and product recommendation text pre-stored in the server can refer to the foregoing embodiments, and the embodiments of this application will not be described in detail here.
示例性地，图10为本申请示例性实施例提供的一种商品推荐文本显示页面的示意图。如图10所示出的，在客户端的商品推荐页面中，客户端显示有针对第二用户推荐的目标商品“陶瓷茶杯”和“小风扇”，针对每个目标商品，客户端同时显示有目标商品推荐文本。针对目标商品“陶瓷茶杯”，客户端显示的目标商品推荐文本1001为“3000+浙江用户购买”；针对目标商品“小风扇”，客户端显示的目标商品推荐文本1002为“8000+上班族购买”。这样，客户端基于与第二用户相同人群标签的用户行为来进行推荐文本的生成，符合第二用户的人群特征，更能够契合第二用户的实际需求。Exemplarily, FIG. 10 is a schematic diagram of a product recommendation text display page provided by an exemplary embodiment of the present application. As shown in Figure 10, in the product recommendation page of the client, the client displays the target products "ceramic tea cup" and "small fan" recommended for the second user. For each target product, the client also displays the target product Product recommendation text. For the target product "ceramic tea cup", the target product recommendation text 1001 displayed on the client is "3000+ Zhejiang users purchased"; for the target product "small fan", the target product recommendation text 1002 displayed on the client is "8000+ office workers purchased" ". In this way, the client generates recommended text based on user behavior with the same crowd tag as the second user, which is consistent with the crowd characteristics of the second user and can better meet the actual needs of the second user.
在上述实施例的基础上，图11为本申请示例性实施例提供的一种商品推荐文本生成方法的流程示意图。请参见图11，该商品推荐文本生成方法可以包括：Based on the above embodiments, FIG. 11 is a schematic flowchart of a method for generating product recommendation text provided by an exemplary embodiment of the present application. Please refer to Figure 11. The product recommendation text generation method may include:
S1101、获取第二用户的用户标识，并根据用户标识，确定第二用户对应的人群标识以及目标商品标识。S1101. Obtain the user identification of the second user, and determine the group identification and target product identification corresponding to the second user based on the user identification.
S1102、基于人群标识以及目标商品标识，从商品推荐文本中确定第二用户对应的目标商品推荐文本。S1102. Based on the crowd identification and the target product identification, determine the target product recommendation text corresponding to the second user from the product recommendation text.
S1103、向客户端发送目标商品标识以及目标商品推荐文本。S1103. Send the target product identification and target product recommendation text to the client.
本申请实施例的执行主体可以为服务端，例如服务器等，也可以为设置在服务端中的商品推荐文本生成装置。商品推荐文本生成装置可以通过软件实现，也可以通过软件和硬件的结合实现。为了便于理解，在下文中，以执行主体为服务端为例进行说明。The execution subject of the embodiment of the present application may be a server, such as a server, or may be a product recommendation text generating device provided in the server. The product recommendation text generating device can be implemented by software or a combination of software and hardware. For ease of understanding, in the following description, the execution subject is the server as an example.
本申请实施例中，服务端可以获取客户端上传的用户标识，然后基于该用户标识，确定出第二用户对应的人群标识，并且可以根据商品推荐算法确定出第二用户对应的目标商品标识。之后，服务端可以基于人群标识、目标商品标识，可以从预先存储的目标数据表中确定出第二用户对应的目标商品推荐文本。在确定出目标商品标识以及目标商品推荐文本之后，服务端可以将目标商品标识以及目标商品推荐文本返回至客户端，由客户端在商品推荐页面中进行显示。这样，服务端可以基于预先存储商品标识-人群标识-商品推荐文本的目标数据表，基于第二用户的用户标识快速确定出待推荐的目标商品标识以及目标商品推荐文本，使得商品推荐更加契合第二用户的实际需求，能够提高对消费者的吸引力。In the embodiment of this application, the server can obtain the user ID uploaded by the client, and then determine the group ID corresponding to the second user based on the user ID, and determine the target product ID corresponding to the second user based on the product recommendation algorithm. Afterwards, the server can determine the target product recommendation text corresponding to the second user from the pre-stored target data table based on the crowd identification and the target product identification. After determining the target product identification and target product recommendation text, the server may return the target product identification and target product recommendation text to the client, and the client may display the target product identification and target product recommendation text on the product recommendation page. In this way, the server can quickly determine the target product identification and target product recommendation text to be recommended based on the user ID of the second user based on the target data table that pre-stores the product identification-crowd identification-product recommendation text, so that the product recommendation is more suitable for the second user. Second, the actual needs of users can improve the appeal to consumers.
需要说明的，服务端中预先存储的人群标识、商品推荐文本的确定过程可以参照前述实施例，本申请实施例在此不再赘述。应理解，在本申请的各种实施例中，各过程的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本申请实施例的实施过程构成任何限定。It should be noted that the process of determining the crowd identification and product recommendation text pre-stored in the server can refer to the foregoing embodiments, and the embodiments of this application will not be described in detail here. It should be understood that in the various embodiments of the present application, the size of the sequence numbers of each process does not mean the order of execution. The execution order of each process should be determined by its functions and internal logic, and should not be determined by the execution order of the embodiments of the present application. The implementation process constitutes no limitation.
在上述任意一个实施例的基础上，图12为本申请示例性实施例提供的一种商品推荐文本配置的技术链路示意图。如图12所示出的，第一用户对应的客户端(第一客户端)响应于第一用户的配置操作，确定出第一关联数据、第二关联数据以及目标格式文本。之后第一客户端将第一关联数据、第二关联数据以及目标格式文本这些配置信息发送至服务端进行执行，服务端基于第二关联数据可以创建人群标识，并确定各个人群标识包括的用户标识；服务端通过Dag调度获取历史数据，并按照目标格式文本来统计目标行为数据，进行字段拼接得到商品推荐文本。服务端最终可以将商品标识-人群标识-商品推荐文本关联存储至目标数据表中。服务端可以响应于第一用户的数据导出操作，生成目标文件并向第一客户端发送该目标文件。Based on any of the above embodiments, FIG. 12 is a schematic diagram of technical links for product recommendation text configuration provided by an exemplary embodiment of the present application. As shown in Figure 12, the client corresponding to the first user (first client) determines the first associated data, the second associated data and the target format text in response to the first user's configuration operation. Afterwards, the first client sends the configuration information of the first associated data, the second associated data and the target format text to the server for execution. The server can create a crowd identifier based on the second associated data and determine the user identifier included in each crowd identifier. ; The server obtains historical data through Dag scheduling, counts target behavior data according to the target format text, and performs field splicing to obtain product recommendation text. The server can finally store the product identification-crowd identification-product recommendation text association into the target data table. The server may generate a target file and send the target file to the first client in response to the first user's data export operation.
第二用户对应的客户端(第二客户端)响应于第二用户的页面加载操作，可以确定出第二用户的用户标识并向服务端发送该用户标识，服务端可以基于该用户标识确定出第二用户对应的人群标识，并确定出第二用户对应目标商品标识；然后服务端可以基于第二用户的人群标识、目标商品标识，从目标数据表中确定出第二用户对应的目标商品推荐文本，之后可以将目标商品标识以及目标商品推荐文本发送至第二客户端，第二客户端基于目标商品标识，可以获取到目标商品的具体信息，例如商品图像、标题、价格信息、评论信以及、型号信息等，之后可以在商品推荐页面中进行展示目标商品以及目标商品推荐文本。In response to the second user's page loading operation, the client corresponding to the second user (second client) can determine the user ID of the second user and send the user ID to the server. The server can determine the user ID based on the user ID. The crowd identification corresponding to the second user, and determine the target product identification corresponding to the second user; then the server can determine the target product recommendation corresponding to the second user from the target data table based on the crowd identification and target product identification of the second user text, and then the target product identification and target product recommendation text can be sent to the second client. Based on the target product identification, the second client can obtain specific information of the target product, such as product images, titles, price information, comment letters, and , model information, etc., and then the target product and target product recommendation text can be displayed on the product recommendation page.
在本申请实施例中，服务端基于第一用户配置的第二关联数据，可以自动快速批量创建多个人群标识，无需第一用户逐个进行手动创建，提高了操作效率，能够减少时间成本，同时可以避免多次人工操作出错率较高的情况。服务端结合消费者的历史数据，对消费者的行为数据进行统计，结合商品标签等文本元素生成针对人群标识的商品推荐文本，能够为商品生成人群维度的符合群体属性的商品推荐文本，更加契合第二用户的实际需求。服务端基于Dag任务调度，可以检测第一关联数据以及第二关联数据的更新状态，实现商品推荐文本的动态更新，能够确保商品推荐文本的时效性；同时也能够实现目标文件的一键导出，可以简化第一用户的操作，节省时间成本。In the embodiment of this application, the server can automatically and quickly create multiple crowd identifiers in batches based on the second associated data configured by the first user, without the need for the first user to manually create one by one, which improves operational efficiency and can reduce time costs. At the same time, This can avoid multiple manual operations with a high error rate. The server combines the historical data of consumers to collect statistics on consumer behavior data, and combines text elements such as product tags to generate product recommendation text targeted at crowd identification. It can generate product recommendation text for products in the crowd dimension that conforms to group attributes, and is more suitable for The actual needs of the second user. Based on Dag task scheduling, the server can detect the update status of the first associated data and the second associated data, realize dynamic updating of product recommendation text, and ensure the timeliness of product recommendation text; it can also realize one-click export of target files. It can simplify the operation of the first user and save time and cost.
图13为本申请示例性实施例提供的一种商品推荐文本生成装置的结构示意图，请参见图13，商品推荐文本生成装置130包括：Figure 13 is a schematic structural diagram of a product recommendation text generation device provided by an exemplary embodiment of the present application. Please refer to Figure 13. The product recommendation text generation device 130 includes:
第一获取模块131，用于获取第一关联数据、第二关联数据以及目标格式文本；第一关联数据包括商品标识以及商品标签；第二关联数据包括用户标识以及人群标签；目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；The first acquisition module 131 is used to obtain the first associated data, the second associated data and the target format text; the first associated data includes product identification and product tags; the second associated data includes user identification and group tags; the target format text includes pre- Assume that there are at least two types of description text, product label, time period and behavioral data;
创建模块132，用于根据第二关联数据，创建人群标签对应的人群标识并确定人群标识包括的用户标识；The creation module 132 is configured to create a crowd identification corresponding to the crowd tag and determine the user ID included in the crowd identification according to the second associated data;
第二获取模块133，用于在目标格式文本中包括行为数据的情况下，获取商品标识在时间周期内对应的历史数据，并根据历史数据、人群标识以及行为数据，确定人群标识对应的目标行为数据；The second acquisition module 133 is used to obtain the historical data corresponding to the product identification within the time period when the target format text includes behavioral data, and determine the target behavior corresponding to the crowd identification based on the historical data, crowd identification and behavioral data. data;
生成模块134，用于基于目标行为数据，按照目标格式文本生成商品标识对应的商品推荐文本。The generation module 134 is configured to generate product recommendation text corresponding to the product identification according to the target format text based on the target behavior data.
在一种可能的实施方式中，创建模块132，具体用于：In a possible implementation, the creation module 132 is specifically used for:
读取第二关联数据中的多个人群标签，分别创建各个人群标签对应的人群标识；Read multiple crowd tags in the second associated data, and create crowd identifiers corresponding to each crowd tag;
根据第二关联数据中用户标识与人群标签的关联关系，确定各个人群标识包括的用户标识；Determine the user identifiers included in each crowd identifier according to the association between the user identifiers and the crowd tags in the second associated data;
存储人群标识以及人群标识包括的用户标识。Store crowd identification and user IDs included in the crowd identification.
在一种可能的实施方式中，第二获取模块133，具体用于：In a possible implementation, the second acquisition module 133 is specifically used to:
周期性检测第一关联数据与第二关联数据的更新状态；Periodically detect the update status of the first associated data and the second associated data;
在第一关联数据与第二关联数据更新完成后，获取商品标识在时间周期内的历史数据；历史数据包括商品标识对应的历史行为数据。After the first associated data and the second associated data are updated, historical data of the product identification within the time period is obtained; the historical data includes historical behavior data corresponding to the product identification.
在一种可能的实施方式中，第二获取模块133，具体用于：In a possible implementation, the second acquisition module 133 is specifically used to:
确定行为数据对应的目标行为类型；Determine the target behavior type corresponding to the behavioral data;
在历史数据中，确定人群标识对应的目标行为类型的数量，得到人群标识对应的目标行为数据。In the historical data, the number of target behavior types corresponding to the crowd identification is determined, and the target behavior data corresponding to the crowd identification is obtained.
在一种可能的实施方式中，装置130还用于：In a possible implementation, the device 130 is also used for:
基于商品标识、人群标识以及商品推荐文本之间的关联关系，将商品标识、人群标识以及商品推荐文本存储至目标数据表中；Based on the association between product identification, crowd identification and product recommendation text, store the product identification, crowd identification and product recommendation text into the target data table;
接收客户端发送的数据导出请求，生成目标文件并向客户端发送目标文件。Receive the data export request sent by the client, generate the target file and send the target file to the client.
在一种可能的实施方式中，商品标签包括地理位置标签、季节时令标签、身份标签、类别标签其中至少一种。In a possible implementation, the product tag includes at least one of a geographical location tag, a seasonal tag, an identity tag, and a category tag.
本申请实施例提供的商品推荐文本生成装置130可以执行上述方法实施例所示的技术方案，其实现原理以及有益效果类似，此处不再进行赘述。The product recommendation text generating device 130 provided by the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be described again here.
图14为本申请示例性实施例提供的一种商品推荐文本生成装置的结构示意图，请参见图14，商品推荐文本生成装置140包括：Figure 14 is a schematic structural diagram of a product recommendation text generation device provided by an exemplary embodiment of the present application. Please refer to Figure 14. The product recommendation text generation device 140 includes:
第一确定模块141，用于响应于第一用户的第一触控操作，确定第一关联数据以及第二关联数据；第一关联数据包括商品标识以及商品标签；第二关联数据包括用户标识以及人群标签；The first determination module 141 is configured to determine first associated data and second associated data in response to the first user's first touch operation; the first associated data includes product identification and product tags; the second associated data includes user identification and crowd tag;
第二确定模块142，用于响应于第一用户的第二触控操作，确定目标格式文本；目标格式文本包括预设描述文本、商品标签、时间周期以及行为数据其中至少两种；The second determination module 142 is configured to determine the target format text in response to the first user's second touch operation; the target format text includes at least two of preset description text, product tags, time period and behavior data;
发送模块143，用于向服务端发送第一关联数据、第二关联数据以及目标格式文本，以使得服务端基于第二关联数据创建人群标签对应的人群标识并确定人群标识包括的用户标识，并且基于目标格式文本生成商品标识对应的商品推荐文本。The sending module 143 is configured to send the first associated data, the second associated data and the target format text to the server, so that the server creates a crowd identifier corresponding to the crowd tag based on the second associated data and determines the user identifier included in the crowd identifier, and Generate product recommendation text corresponding to the product identification based on the target format text.
在一种可能的实施方式中，第二确定模块142，具体用于：In a possible implementation, the second determination module 142 is specifically used to:
响应于第二触控操作，基于第二触控操作对应的文本元素生成目标格式文本；或者，In response to the second touch operation, generate target format text based on the text element corresponding to the second touch operation; or,
响应于第二触控操作，基于第二触控操作对应的文本元素生成目标格式文本，并确定目标格式文本对应的其他历史数据以及查询条件。In response to the second touch operation, a target format text is generated based on the text element corresponding to the second touch operation, and other historical data and query conditions corresponding to the target format text are determined.
在一种可能的实施方式中，装置140还用于：In a possible implementation, the device 140 is also used for:
响应于第一用户的数据导出操作，向服务端发送数据导出请求，以使得服务端基于数据导出请求生成目标文件；目标文件中包括具备关联关系的商品标识、人群标识以及商品推荐文本；In response to the first user's data export operation, a data export request is sent to the server, so that the server generates a target file based on the data export request; the target file includes associated product identifiers, group identifiers, and product recommendation text;
接收服务端返回的目标文件。Receive the target file returned by the server.
本申请实施例提供的商品推荐文本生成装置140可以执行上述方法实施例所示的技术方案，其实现原理以及有益效果类似，此处不再进行赘述。The product recommendation text generation device 140 provided by the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be described again here.
图15为本申请示例性实施例提供的一种商品推荐文本生成装置的结构示意图，请参见图15，商品推荐文本生成装置150包括：Figure 15 is a schematic structural diagram of a product recommendation text generation device provided by an exemplary embodiment of the present application. Please refer to Figure 15. The product recommendation text generation device 150 includes:
确定模块151，用于响应于第二用户的页面加载操作，确定第二用户的用户标识；Determining module 151, configured to determine the user identification of the second user in response to the second user's page loading operation;
发送模块152，用于向服务端发送用户标识，以使得服务端根据用户标识确定第二用户对应的人群标识以及目标商品标识，并基于人群标识以及目标商品标识从商品推荐文本中确定第二用户对应的目标商品推荐文本；The sending module 152 is used to send the user identification to the server, so that the server determines the crowd identification and the target product identification corresponding to the second user based on the user identification, and determines the second user from the product recommendation text based on the crowd identification and the target product identification. Corresponding target product recommendation text;
展示模块153，用于展示目标商品标识对应的目标商品以及目标商品推荐文本。The display module 153 is used to display the target product corresponding to the target product identification and the target product recommendation text.
本申请实施例提供的商品推荐文本生成装置150可以执行上述方法实施例所示的技术方案，其实现原理以及有益效果类似，此处不再进行赘述。The product recommendation text generating device 150 provided by the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be described again here.
图16为本申请示例性实施例提供的一种商品推荐文本生成装置的结构示意图，请参见图16，商品推荐文本生成装置160包括：Figure 16 is a schematic structural diagram of a product recommendation text generation device provided by an exemplary embodiment of the present application. Please refer to Figure 16. The product recommendation text generation device 160 includes:
获取模块161，用于获取第二用户的用户标识，并根据用户标识，确定第二用户对应的人群标识以及目标商品标识；The acquisition module 161 is used to obtain the user identification of the second user, and determine the group identification and target product identification corresponding to the second user based on the user identification;
确定模块162，用于基于人群标识以及目标商品标识，从商品推荐文本中确定第二用户对应的目标商品推荐文本；The determination module 162 is configured to determine the target product recommendation text corresponding to the second user from the product recommendation text based on the crowd identification and the target product identification;
发送模块163，用于向客户端发送目标商品标识以及目标商品推荐文本。The sending module 163 is used to send the target product identification and the target product recommendation text to the client.
本申请实施例提供的商品推荐文本生成装置160可以执行上述方法实施例所示的技术方案，其实现原理以及有益效果类似，此处不再进行赘述。The product recommendation text generation device 160 provided by the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be described again here.
图17为本申请示例性实施例提供的一种电子设备的结构示意图，请参见图17，该电子设备170可以包括处理器171和存储器172。示例性地，处理器171、存储器172，各部分之间通过总线173相互连接。FIG. 17 is a schematic structural diagram of an electronic device according to an exemplary embodiment of the present application. Please refer to FIG. 17 . The electronic device 170 may include a processor 171 and a memory 172 . For example, the processor 171 and the memory 172 are connected to each other through a bus 173 .
存储器172存储计算机执行指令；Memory 172 stores computer execution instructions;
处理器171执行存储器172存储的计算机执行指令，使得处理器171执行如上述方法实施例所示的商品推荐文本生成方法。The processor 171 executes the computer execution instructions stored in the memory 172, so that the processor 171 executes the product recommendation text generation method as shown in the above method embodiment.
相应地，本申请实施例提供一种计算机可读存储介质，计算机可读存储介质中存储有计算机执行指令，当计算机执行指令被处理器执行时用于实现上述方法实施例的商品推荐文本生成方法。Correspondingly, embodiments of the present application provide a computer-readable storage medium. Computer-executable instructions are stored in the computer-readable storage medium. When the computer-executable instructions are executed by a processor, they are used to implement the product recommendation text generation method of the above method embodiment. .
相应地，本申请实施例还可提供一种计算机程序产品，包括计算机程序，该计算机程序被处理器执行时，可实现上述方法实施例所示的商品推荐文本生成方法。Correspondingly, embodiments of the present application can also provide a computer program product, including a computer program. When the computer program is executed by a processor, it can implement the product recommendation text generation method shown in the above method embodiment.
本领域内的技术人员应明白，本发明的实施例可提供为方法、系统、或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。The invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each process and/or block in the flowchart illustrations and/or block diagrams, and combinations of processes and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing device produce a use A device for realizing the functions specified in one process or multiple processes of the flowchart and/or one block or multiple blocks of the block diagram.
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。These computer program instructions may also be stored in a computer-readable memory that causes a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction means, the instructions The device implements the functions specified in a process or processes of the flowchart and/or a block or blocks of the block diagram.
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。These computer program instructions may also be loaded onto a computer or other programmable data processing device, causing a series of operating steps to be performed on the computer or other programmable device to produce computer-implemented processing, thereby executing on the computer or other programmable device. Instructions provide steps for implementing the functions specified in a process or processes of a flowchart diagram and/or a block or blocks of a block diagram.
在一个典型的配置中，计算设备包括一个或多个处理器、输入/输出接口、网络接口和内存。In a typical configuration, a computing device includes one or more processors, input/output interfaces, network interfaces, and memory.
内存可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)和/或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。Memory may include non-permanent storage in computer-readable media, random access memory (RAM), and/or non-volatile memory in the form of read-only memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。Computer-readable media includes both persistent and non-volatile, removable and non-removable media that can be implemented by any method or technology for storage of information. Information may be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, Magnetic tape cassettes, tape disk storage or other magnetic storage devices or any other non-transmission medium can be used to store information that can be accessed by a computing device. As defined in this article, computer-readable media does not include transitory media, such as modulated data signals and carrier waves.
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、商品或者设备中还存在另外的相同要素。It should also be noted that the terms "comprises," "comprises," or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements not only includes those elements, but also includes Other elements are not expressly listed or are inherent to the process, method, article or equipment. Without further limitation, an element defined by the statement "comprises a..." does not exclude the presence of additional identical elements in a process, method, article, or device that includes the stated element.
以上所述仅为本申请的实施例而已，并不用于限制本申请。对于本领域技术人员来说，本申请可以有各种更改和变化。凡在本申请的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本申请的权利要求范围之内。The above descriptions are only examples of the present application and are not intended to limit the present application. To those skilled in the art, various modifications and variations may be made to this application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application shall be included in the scope of the claims of this application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310878110.7ACN116883111A (en) | 2023-07-17 | 2023-07-17 | Product recommendation text generation method, device and equipment |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310878110.7ACN116883111A (en) | 2023-07-17 | 2023-07-17 | Product recommendation text generation method, device and equipment |
| Publication Number | Publication Date |
|---|---|
| CN116883111Atrue CN116883111A (en) | 2023-10-13 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310878110.7APendingCN116883111A (en) | 2023-07-17 | 2023-07-17 | Product recommendation text generation method, device and equipment |
| Country | Link |
|---|---|
| CN (1) | CN116883111A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117808564A (en)* | 2024-03-01 | 2024-04-02 | 深圳市拓保软件有限公司 | User data recommendation method and device based on artificial intelligence |
| CN119205254A (en)* | 2024-09-02 | 2024-12-27 | 杭州衣科信息技术股份有限公司 | Product information publishing method, system and device |
| WO2025106019A1 (en)* | 2023-11-16 | 2025-05-22 | Grabtaxi Holdings Pte. Ltd. | Persona-based app content personalisation |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2025106019A1 (en)* | 2023-11-16 | 2025-05-22 | Grabtaxi Holdings Pte. Ltd. | Persona-based app content personalisation |
| CN117808564A (en)* | 2024-03-01 | 2024-04-02 | 深圳市拓保软件有限公司 | User data recommendation method and device based on artificial intelligence |
| CN117808564B (en)* | 2024-03-01 | 2024-05-28 | 深圳市拓保软件有限公司 | User data recommendation method and device based on artificial intelligence |
| CN119205254A (en)* | 2024-09-02 | 2024-12-27 | 杭州衣科信息技术股份有限公司 | Product information publishing method, system and device |
| CN119205254B (en)* | 2024-09-02 | 2025-10-03 | 杭州衣科信息技术股份有限公司 | Product information publishing method, system and device |
| Publication | Publication Date | Title |
|---|---|---|
| JP7618079B2 (en) | Actionable Widget Cards | |
| CN112559896B (en) | Information recommendation method, device, equipment and computer readable storage medium | |
| US10482490B2 (en) | Behavioral tracking system and method in support of high-engagement communications | |
| US9495661B2 (en) | Embeddable context sensitive chat system | |
| CN111164950B (en) | Method and system for user segmentation as a service | |
| US9984386B1 (en) | Rules recommendation based on customer feedback | |
| CN116883111A (en) | Product recommendation text generation method, device and equipment | |
| US20170236160A1 (en) | System and method for specifying targeted content for customers | |
| US20140279016A1 (en) | Behavioral tracking system and method in support of high-engagement communications | |
| CA2907874A1 (en) | On-site and in-store content personalization and optimization | |
| WO2013043929A1 (en) | Method and apparatus for the development, delivery and deployment of action-oriented business applications supported by a cloud based action server platform | |
| US20210150604A1 (en) | Systems and methods for customization of reviews | |
| GB2518936A (en) | Method of optimization for an application | |
| US20210011957A1 (en) | Web page designing with content beam to device preview | |
| US9442745B2 (en) | Business-to-consumer extendable base application | |
| US20250097349A1 (en) | Simultaneous Voice and Data Content Driven Commercial Data Platform | |
| US20210150593A1 (en) | Systems and methods for customization of reviews | |
| US20140289054A1 (en) | Behavioral tracking system and method in support of high-engagement communications | |
| CN112785324A (en) | Information processing method, information processing apparatus, electronic device, and medium | |
| US20240124403A1 (en) | Actionable widget cards | |
| Safonov et al. | Digitalisation as a marketing management tool in the tourism industry | |
| US20200211082A1 (en) | Methods and system for product frame guided producer-buyer platform interactions | |
| AU2015100778B4 (en) | Smart lists | |
| US20240220710A1 (en) | System and method for integrated temporal external resource allocation | |
| CN113806626B (en) | A method and system for sending push messages |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |