技术领域technical field
本发明涉及数据安全技术领域，尤其涉及一种基于知识图谱的敏感识别方法。The invention relates to the technical field of data security, in particular to a sensitive identification method based on a knowledge graph.
背景技术Background technique
推荐系统是一种信息过滤工具，它旨在准确地预测用户对商品的偏好程度，从而把对用户更有价值的商品优先呈现给他们。而用户历史行为数据是推荐系统的支撑基础，用户的历史行为数据常常涉及到用户的个人敏感数据。对敏感数据进行隐私保护的前提是能从大量的数据中挑选出敏感数据，完成对敏感数据的识别。The recommendation system is an information filtering tool, which aims to accurately predict the user's preference for the product, so as to present the product that is more valuable to the user to them first. The user's historical behavior data is the supporting basis of the recommendation system, and the user's historical behavior data often involves the user's personal sensitive data. The premise of privacy protection for sensitive data is to be able to select sensitive data from a large amount of data and complete the identification of sensitive data.
传统的敏感数据的识别方法主要有字典匹配方法和人工识别这两种。业界大多采用字典匹配方法和人工识别方法相结合的方式对敏感数据进行识别。主要过程如下：用户定义敏感数据模式匹配式，根据预定义的模型确定字典匹配范围，然后使用字典匹配对目标进行匹配扫描，在完成扫描后，通过人工对匹配结果过滤，并对模式数据匹配式进行优化，但是因评判标准和字典匹配问题会导致识别速度慢。Traditional sensitive data identification methods mainly include dictionary matching method and manual identification. Most of the industry uses a combination of dictionary matching methods and manual identification methods to identify sensitive data. The main process is as follows: the user defines the sensitive data pattern matching formula, determines the dictionary matching range according to the predefined model, and then uses the dictionary matching to scan the target for matching, after the scanning is completed, manually filters the matching results, and matches the pattern data It is optimized, but the recognition speed will be slow due to the judging criteria and dictionary matching problems.
发明内容Contents of the invention
本发明的目的在于提供一种基于知识图谱的敏感识别方法，提高识别速度。The purpose of the present invention is to provide a sensitive recognition method based on a knowledge map, and improve the recognition speed.
为实现上述目的，本发明提供了一种基于知识图谱的敏感识别方法，包括以下步骤：In order to achieve the above purpose, the present invention provides a sensitive identification method based on knowledge graph, comprising the following steps:
对获取的原始数据进行预处理，并构建用户物品的模式图；Preprocess the acquired raw data and build a pattern diagram of user items;
根据所述模式图和预处理后的数据，构建知识图谱；Construct a knowledge map according to the pattern map and the preprocessed data;
构建敏感关系推理规则，并补全所述知识图谱；Construct sensitive relational inference rules and complete the knowledge graph;
对所述知识图谱中的敏感数据进行查询，并输出所述敏感数据。Query sensitive data in the knowledge map, and output the sensitive data.
其中，对获取的原始数据进行预处理，并构建用户物品的模式图，包括：Among them, the acquired raw data is preprocessed, and a pattern map of user items is constructed, including:
将获取的多种类型的原始数据中的数据存储格式和编码方法进行统一，同时对冗余的数据进行删除。Unify the data storage format and encoding method in the various types of raw data obtained, and delete redundant data at the same time.
其中，对获取的原始数据进行预处理，并构建用户物品的模式图，还包括：Among them, the acquired raw data is preprocessed, and a pattern diagram of user items is constructed, which also includes:
将用户年龄、职业、性别作为用户的属性，并标记用户与物品间的关系为购买关系，然后采用数据库工具对预处理后的数据进行实体对齐，构建用户物品的模式图。The user's age, occupation, and gender are used as the attributes of the user, and the relationship between the user and the item is marked as the purchase relationship, and then the database tool is used to perform entity alignment on the preprocessed data to build a schema diagram of the user's item.
其中，根据所述模式图和预处理后的数据，构建知识图谱，包括：Wherein, according to the pattern diagram and the preprocessed data, a knowledge map is constructed, including:
将用户和物品作为节点，并根据获取的所述用户和物品的每一个属性的键值对构建属性图模型。Taking users and items as nodes, and constructing an attribute graph model according to the acquired key-value pairs of each attribute of the users and items.
其中，根据所述模式图和预处理后的数据，构建知识图谱，还包括：Wherein, constructing a knowledge map according to the pattern map and preprocessed data also includes:
将所述用户映射为头实体，将所述物品映射为尾实体，同时将所述用户与对应的所述物品之间的关系映射为0或1，并采用图数据库存储知识图谱。The user is mapped to the head entity, the item is mapped to the tail entity, and the relationship between the user and the corresponding item is mapped to 0 or 1, and a graph database is used to store the knowledge graph.
其中，对所述知识图谱中的敏感数据进行查询，并输出所述敏感数据，包括：Wherein, the sensitive data in the knowledge map is queried, and the sensitive data is output, including:
利用图形查询语言查询补全后的所述知识图谱中的图形数据，根据声明的查询目标，返回所有具有对应敏感关系的用户和物品节点。Using a graph query language to query the graph data in the completed knowledge graph, and return all user and item nodes with corresponding sensitive relationships according to the declared query target.
其中，对所述知识图谱中的敏感数据进行查询，并输出所述敏感数据，还包括：Wherein, querying the sensitive data in the knowledge graph and outputting the sensitive data also includes:
根据数据存储格式和编码方法，将返回的敏感节点还原为对应的所述原始数据，并存入对应的保存文件。According to the data storage format and encoding method, the returned sensitive nodes are restored to the corresponding original data, and stored in the corresponding saved file.
本发明的一种基于知识图谱的敏感识别方法，首先为了构建用户-商品知识图谱，需要对获取的原始数据集进行预处理，并通过预处理后的数据构建用户-物品的模式图，然后根据预处理后的数据和所述模式图构建知识图谱；其次，为了识别敏感数据，通过构建的敏感关系推理规则补全知识图谱中用户与物品间原本不存在的敏感关系；最后，对整个所述知识图谱查询出敏感数据，并输出，提高识别速度。In the sensitive recognition method based on knowledge graph of the present invention, firstly, in order to construct the user-product knowledge graph, it is necessary to preprocess the acquired original data set, and construct a user-item pattern diagram through the preprocessed data, and then according to The preprocessed data and the pattern map construct a knowledge map; secondly, in order to identify sensitive data, the sensitive relationship between the user and the item in the knowledge map is complemented by the constructed sensitive relationship reasoning rules; finally, the entire said The knowledge map queries sensitive data and outputs it to improve the recognition speed.
附图说明Description of drawings
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.
图1是本发明提供的一种基于知识图谱的敏感识别方法的步骤示意图。Fig. 1 is a schematic diagram of the steps of a sensitive identification method based on a knowledge map provided by the present invention.
图2是本发明提供的一种基于知识图谱的敏感识别方法的流程示意图。Fig. 2 is a schematic flowchart of a sensitive identification method based on a knowledge map provided by the present invention.
图3是本发明提供的用户-商品的模式图。Fig. 3 is a schematic diagram of user-commodity provided by the present invention.
图4是本发明提供的属性图模型。Fig. 4 is an attribute graph model provided by the present invention.
具体实施方式Detailed ways
下面详细描述本发明的实施例，所述实施例的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施例是示例性的，旨在用于解释本发明，而不能理解为对本发明的限制。Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals designate the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.
在本发明的描述中，“多个”的含义是两个或两个以上，除非另有明确具体的限定。In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.
请参阅图1和图2，本发明提供一种基于知识图谱的敏感识别方法，包括以下步骤：Please refer to Fig. 1 and Fig. 2, the present invention provides a kind of sensitive identification method based on knowledge graph, comprises the following steps:
S101、对获取的原始数据进行预处理，并构建用户物品的模式图。S101. Perform preprocessing on the acquired raw data, and construct a model diagram of the user's item.
具体的，原始的数据包括结构化的评分数据和非结构化的文本数据，如用户的评论、产品的描述，该部分工作主要包括以下几个方面：Specifically, the original data includes structured rating data and unstructured text data, such as user reviews and product descriptions. This part of the work mainly includes the following aspects:
1、对于多种类型的数据，先统一数据存储格式和编码方法，为了满足后续的知识抽取和数据存储的要求，这里将多数据源的评分数据转换为包含用户ID、物品ID、用户属性、物品属性的文件。1. For multiple types of data, first unify the data storage format and encoding method. In order to meet the subsequent knowledge extraction and data storage requirements, here the scoring data of multiple data sources is converted to include user ID, item ID, user attribute, A file of item properties.
2、由于电子商务数据中存在很多冗余的数据，如重复的评分记录、文本中的广告信息、重复的段落以及质量低的数据，这里对这些可信度低、不完整的数据进行去除，使数据规范化、提高数据的可用性，同时也可以减少后续的数据识别量，提高识别速度。2. Since there are many redundant data in e-commerce data, such as repeated scoring records, advertisement information in text, repeated paragraphs and low-quality data, these low-reliability and incomplete data are removed here. Standardize the data, improve the availability of data, and also reduce the amount of subsequent data recognition and improve the speed of recognition.
经过预处理后得到的数据主要有ratings.dat、users.dat、product.dat三个文件，其中ratings.dat中主要存储的是用户及其对物品的评分，users.dat主要存储的是用户及其属性(性别、年龄、职业等)，product.dat主要存储的是物品及其属性(如类别)。The data obtained after preprocessing mainly include three files: ratings.dat, users.dat, and product.dat. Among them, ratings.dat mainly stores users and their ratings on items, and users.dat mainly stores users and Its attributes (gender, age, occupation, etc.), product.dat mainly stores items and their attributes (such as categories).
选择用户年龄、职业、性别作为用户的属性，标记用户与物品间的关系为购买关系。采用数据库工具对数据集中的数据进行实体对齐。通过预处理后的信息构建用户物品的模式图，如图3所示，包含用户、商品、商品的相关属性、用户的相关属性及用户与商品间的购买关系。Select the user's age, occupation, and gender as the attributes of the user, and mark the relationship between the user and the item as the purchase relationship. Use database tools to perform entity alignment on the data in the dataset. The model diagram of user items is constructed through the preprocessed information, as shown in Figure 3, including users, products, related attributes of products, related attributes of users, and the purchase relationship between users and products.
S102、根据所述模式图和预处理后的数据，构建知识图谱。S102. Construct a knowledge map according to the pattern map and the preprocessed data.
具体的，通过预处理后的数据及设计好的模式图构建用户商品知识图谱。主要包括知识表示、知识抽取、知识融合等步骤。Specifically, the user commodity knowledge map is constructed through the preprocessed data and the designed pattern map. It mainly includes knowledge representation, knowledge extraction, knowledge fusion and other steps.
知识表示的具体步骤为：The specific steps of knowledge representation are:
用属性图模型作为数据模型。用户和物品表示为节点，用户节点具有年龄、职业、性别等属性，物品节点具有类别、价格等属性，每个属性是一个键值对。每条边具有一个标签，表示联系。边也同样具有属性。例如，“用户李明曾经购买了一支笔......”对应的属性如图4所示，对应的三元组可以表示为(李明，购买，笔)，意思为头实体“李明”和尾实体“笔”间存在“购买”的联系，同时还显示用户的年龄和商品的价格。Use the property graph model as the data model. Users and items are represented as nodes. User nodes have attributes such as age, occupation, and gender. Item nodes have attributes such as category and price. Each attribute is a key-value pair. Each edge has a label, denoting the connection. Edges also have properties. For example, the corresponding attribute of "the user Li Ming once purchased a pen..." is shown in Figure 4, and the corresponding triple can be expressed as (Li Ming, purchase, pen), which means the head entity " There is a "purchase" connection between "Li Ming" and the tail entity "pen", and it also displays the user's age and the price of the product.
知识抽取的具体步骤为：The specific steps of knowledge extraction are:
知识抽取的目的是对结构化数据和非结构化数据进行实体抽取、实体关系的抽取、属性的抽取并存入知识图谱中。对于结构化的数据，直接将用户ID映射为头实体的ID，物品ID映射为尾实体的ID，将用户ID和物品ID间的关系映射为1或0，表示用户与物品间是否进行了交互。如三元组(231,1,324)表示ID为231的用户购买了ID为324的物品。The purpose of knowledge extraction is to extract entities, entity relationships, and attributes from structured and unstructured data and store them in the knowledge graph. For structured data, the user ID is directly mapped to the ID of the head entity, the item ID is mapped to the ID of the tail entity, and the relationship between the user ID and the item ID is mapped to 1 or 0, indicating whether the user interacts with the item . For example, the triplet (231,1,324) indicates that the user with ID 231 purchased the item with ID 324.
存储知识图谱的具体步骤为：The specific steps for storing the knowledge graph are:
用户-物品关系知识图谱是图结构的，采用关系型数据库存储将不利于后续的敏感关系查找和处理，因此，这里采用主流的图数据库Neo4j存储用户-物品知识图谱，将用户实体存储为图数据库中的节点，用户与物品间的关系存储为连接节点的边。The user-item relationship knowledge graph is graph-structured, and using relational database storage will not be conducive to the subsequent search and processing of sensitive relationships. Therefore, the mainstream graph database Neo4j is used here to store the user-item knowledge graph, and the user entity is stored as a graph database. The nodes in , the relationship between users and items are stored as edges connecting the nodes.
经过实体抽取、实体关系抽取和属性抽取后建立了对应的知识图谱，为了对所述知识图谱进行保存，可以对应的图数据库进行数据的保存。本发明采用的图数据库存储的知识图谱，能够很好地融合推荐系统场景中的多源数据，通过图查询语言对数据库进行查询，能够快速地提取敏感数据。After entity extraction, entity relationship extraction, and attribute extraction, a corresponding knowledge graph is established. In order to save the knowledge graph, data can be saved in a corresponding graph database. The knowledge graph stored in the graph database adopted by the present invention can well integrate multi-source data in the recommendation system scene, query the database through graph query language, and quickly extract sensitive data.
S103、构建敏感关系推理规则，并补全所述知识图谱。S103. Construct sensitive relational reasoning rules, and complete the knowledge graph.
具体的，为了识别敏感数据，首先需要构建用户与物品间的敏感关系推理规则。由于用户具有年龄、职业、性别等属性，物品具有价格、类别等属性。定义的规则示例如下：Specifically, in order to identify sensitive data, it is first necessary to construct inference rules for sensitive relationships between users and items. Since users have attributes such as age, occupation, and gender, items have attributes such as price and category. An example of a defined rule is as follows:
Rule1：普通职业的用户经常购买药物，则该物品对当前用户是敏感的。Rule1: If users of ordinary occupations often purchase drugs, the item is sensitive to the current user.
采用Jena工具内置基于规则的推理机进行推理。具体主要包括建模基本模块、构建本体、添加推理机三个步骤。Use the built-in rule-based reasoning engine of the Jena tool for reasoning. Specifically, it mainly includes three steps: modeling basic modules, constructing ontology, and adding reasoning machines.
1、建模所需模块1. Modules required for modeling
首先建立模型最基本的包org.apache.jena.rdf.model用来建立模型。其次建立org.apache.jena.vocabulary.RDF和org.apache.jena.vocabulary.RDFS使用RDF和RDFS中的二元关系。org.apache.jena.reasoner.Reasoner和org.apache.jena.reasoner.ReasonerRegistry用于创建推理机。First build the model The most basic package org.apache.jena.rdf.model is used to build the model. Second, create org.apache.jena.vocabulary.RDF and org.apache.jena.vocabulary.RDFS using the binary relationship in RDF and RDFS. org.apache.jena.reasoner.Reasoner and org.apache.jena.reasoner.ReasonerRegistry are used to create reasoning machines.
2、构建本体2. Build ontology
以上步骤1建立的Model本质上就是Jena中的知识库结构，即本体。The Model established in step 1 above is essentially the knowledge base structure in Jena, that is, ontology.
3、添加推理机3. Add inference engine
直接选用内置的RDFS推理机完成敏感关系推理。Directly select the built-in RDFS reasoning engine to complete sensitive relational reasoning.
执行以上的规则示例Rule1后，知识图谱中所有满足条件的节点间都被添加了敏感关系。After executing the above rule example Rule1, sensitive relationships are added between all nodes that meet the conditions in the knowledge graph.
S104、对所述知识图谱中的敏感数据进行查询，并输出所述敏感数据。S104. Query sensitive data in the knowledge graph, and output the sensitive data.
具体的，由步骤S103，知识图谱中已经补全了用户-物品间的敏感关系。节点间的关系由原来的单一的购买关系扩展为购买关系和敏感关系。这里通过描述性的图形查询语言——Cypher来查询所述知识图谱中的图形数据，主要方法是直接对知识图谱声明“查询的目标”。如以下示例脚本会返回跟sensitive标签有关系的所有节点，即返回所有的具有敏感关系的用户与物品节点。Specifically, by step S103, the sensitive relationship between users and items has been completed in the knowledge graph. The relationship between nodes is expanded from the original single purchase relationship to purchase relationship and sensitive relationship. Here, the graph data in the knowledge graph is queried through Cypher, a descriptive graph query language. The main method is to directly declare the "query target" to the knowledge graph. For example, the following sample script will return all nodes that have a relationship with the sensitive tag, that is, return all user and item nodes that have a sensitive relationship.
match(n)--(m:sensitive)match(n)--(m:sensitive)
returnn；return n;
输出敏感数据output sensitive data
对于识别出的所有敏感节点，根据数据存储格式和编码方法将它们还原为原始的数据，并将他们存入到保存文件中，也可以保存到建立的所述图数据库中，由此就从大量的数据中挑选出了敏感数据，完成了对敏感数据的识别。For all identified sensitive nodes, restore them to the original data according to the data storage format and encoding method, and store them in the saved file, or save them in the established graph database, thus saving a large number of Sensitive data is selected from the data, and the identification of sensitive data is completed.
本发明的创新点包括以下几个方面：The innovations of the present invention include the following aspects:
提出一个个性化的隐私定义，考虑了用户和商品间的敏感性。现有的推荐系统隐私保护问题假设的是用户的反馈数据都是敏感的，而实际上不同用户对不同商品的敏感性是不同的，即不同用户具有不同的隐私保护需求，不同的物品具有不同的敏感程度。A personalized definition of privacy is proposed, considering the sensitivity between users and items. The existing recommendation system privacy protection problem assumes that user feedback data is sensitive, but in fact different users have different sensitivity to different products, that is, different users have different privacy protection needs, and different items have different degree of sensitivity.
提出通过构建用户-商品知识图谱，通过关系推理补全知识图谱中不存在的敏感关系，从而识别出敏感数据，这样能够解决传统敏感数据识别方法的识别速度慢的问题。It is proposed to identify sensitive data by constructing a user-product knowledge graph and completing sensitive relationships that do not exist in the knowledge graph through relational reasoning, which can solve the problem of slow recognition speed of traditional sensitive data recognition methods.
与现有技术相比，本发明方法不仅有效提高敏感数据的识别准确度，而且还能提高面对大量复杂数据时的识别速度。Compared with the prior art, the method of the invention not only effectively improves the recognition accuracy of sensitive data, but also improves the recognition speed when faced with a large amount of complex data.
本发明的一种基于知识图谱的敏感识别方法，首先为了构建用户-商品知识图谱，需要对获取的原始数据集进行预处理，并通过预处理后的数据构建用户-物品的模式图，然后根据预处理后的数据和所述模式图构建知识图谱；其次，为了识别敏感数据，通过构建的敏感关系推理规则补全知识图谱中用户与物品间原本不存在的敏感关系；最后，对整个所述知识图谱查询出敏感数据，并输出，提高识别速度。In the sensitive recognition method based on knowledge graph of the present invention, firstly, in order to construct the user-product knowledge graph, it is necessary to preprocess the acquired original data set, and construct a user-item pattern diagram through the preprocessed data, and then according to The preprocessed data and the pattern map construct a knowledge map; secondly, in order to identify sensitive data, the sensitive relationship between the user and the item in the knowledge map is complemented by the constructed sensitive relationship reasoning rules; finally, the entire said The knowledge map queries sensitive data and outputs it to improve the recognition speed.
以上所揭露的仅为本发明一种较佳实施例而已，当然不能以此来限定本发明之权利范围，本领域普通技术人员可以理解实现上述实施例的全部或部分流程，并依本发明权利要求所作的等同变化，仍属于发明所涵盖的范围。What is disclosed above is only a preferred embodiment of the present invention, and of course it cannot limit the scope of rights of the present invention. Those of ordinary skill in the art can understand all or part of the process for realizing the above embodiments, and according to the rights of the present invention The equivalent changes required still belong to the scope covered by the invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011082927.6ACN112163160B (en) | 2020-10-12 | 2020-10-12 | Sensitive identification method based on knowledge graph |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011082927.6ACN112163160B (en) | 2020-10-12 | 2020-10-12 | Sensitive identification method based on knowledge graph |
| Publication Number | Publication Date |
|---|---|
| CN112163160A CN112163160A (en) | 2021-01-01 |
| CN112163160Btrue CN112163160B (en) | 2023-08-08 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011082927.6AActiveCN112163160B (en) | 2020-10-12 | 2020-10-12 | Sensitive identification method based on knowledge graph |
| Country | Link |
|---|---|
| CN (1) | CN112163160B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113239392A (en)* | 2021-04-02 | 2021-08-10 | 国网福建省电力有限公司信息通信分公司 | Desensitization method based on data center sensitive data |
| CN114021191B (en)* | 2021-11-05 | 2022-07-01 | 江苏安泰信息科技发展有限公司 | Safe production informatization sensitive data management method and system |
| CN114254124A (en)* | 2021-12-16 | 2022-03-29 | 深圳壹账通智能科技有限公司 | Display method, device, computer equipment and medium for sensitive data map |
| CN114186689B (en)* | 2022-02-14 | 2022-05-20 | 支付宝(杭州)信息技术有限公司 | Methods, systems, apparatus, and media for path discovery in a knowledge graph |
| CN114692954A (en)* | 2022-03-09 | 2022-07-01 | 中国科学技术大学 | Service optimization method and platform based on user experience knowledge graph database |
| CN114417845B (en)* | 2022-03-30 | 2022-07-12 | 支付宝(杭州)信息技术有限公司 | Same entity identification method and system based on knowledge graph |
| CN115099909A (en)* | 2022-08-23 | 2022-09-23 | 深圳洽客科技有限公司 | Information processing method and system based on E-commerce intention database mining |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9419841B1 (en)* | 2011-06-29 | 2016-08-16 | Amazon Technologies, Inc. | Token-based secure data management |
| CN108471414A (en)* | 2018-03-24 | 2018-08-31 | 海南大学 | Internet of Things data method for secret protection towards typing resource |
| CN108647532A (en)* | 2018-05-15 | 2018-10-12 | 广东因特利信息科技股份有限公司 | Method, apparatus, electronic equipment and the storage medium of sensitive users mark secrecy |
| CN108804950A (en)* | 2018-06-09 | 2018-11-13 | 海南大学 | Based on data collection of illustrative plates, modeling and the data-privacy guard method of Information Atlas and knowledge mapping |
| CN109284394A (en)* | 2018-09-12 | 2019-01-29 | 青岛大学 | A method for building enterprise knowledge graph from the perspective of multi-source data integration |
| KR101987915B1 (en)* | 2017-12-22 | 2019-06-12 | 주식회사 솔트룩스 | System for generating template used to generate query to knowledge base from natural language question and question answering system including the same |
| CN111259260A (en)* | 2020-03-30 | 2020-06-09 | 九江学院 | A privacy protection method in personalized recommendation based on ranking and classification |
| CN111353091A (en)* | 2018-12-24 | 2020-06-30 | 北京三星通信技术研究有限公司 | Information processing method, apparatus, electronic device, and readable storage medium |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140214895A1 (en)* | 2013-01-31 | 2014-07-31 | Inplore | Systems and method for the privacy-maintaining strategic integration of public and multi-user personal electronic data and history |
| US20150286709A1 (en)* | 2014-04-02 | 2015-10-08 | Samsung Electronics Co., Ltd. | Method and system for retrieving information from knowledge-based assistive network to assist users intent |
| US10542015B2 (en)* | 2016-08-15 | 2020-01-21 | International Business Machines Corporation | Cognitive offense analysis using contextual data and knowledge graphs |
| US10521608B2 (en)* | 2018-01-09 | 2019-12-31 | Accenture Global Solutions Limited | Automated secure identification of personal information |
| US11016965B2 (en)* | 2019-01-22 | 2021-05-25 | International Business Machines Corporation | Graphical user interface for defining atomic query for querying knowledge graph databases |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9419841B1 (en)* | 2011-06-29 | 2016-08-16 | Amazon Technologies, Inc. | Token-based secure data management |
| KR101987915B1 (en)* | 2017-12-22 | 2019-06-12 | 주식회사 솔트룩스 | System for generating template used to generate query to knowledge base from natural language question and question answering system including the same |
| CN108471414A (en)* | 2018-03-24 | 2018-08-31 | 海南大学 | Internet of Things data method for secret protection towards typing resource |
| CN108647532A (en)* | 2018-05-15 | 2018-10-12 | 广东因特利信息科技股份有限公司 | Method, apparatus, electronic equipment and the storage medium of sensitive users mark secrecy |
| CN108804950A (en)* | 2018-06-09 | 2018-11-13 | 海南大学 | Based on data collection of illustrative plates, modeling and the data-privacy guard method of Information Atlas and knowledge mapping |
| CN109284394A (en)* | 2018-09-12 | 2019-01-29 | 青岛大学 | A method for building enterprise knowledge graph from the perspective of multi-source data integration |
| CN111353091A (en)* | 2018-12-24 | 2020-06-30 | 北京三星通信技术研究有限公司 | Information processing method, apparatus, electronic device, and readable storage medium |
| CN111259260A (en)* | 2020-03-30 | 2020-06-09 | 九江学院 | A privacy protection method in personalized recommendation based on ranking and classification |
| Title |
|---|
| 焦佳.社会网络数据发布中一种敏感关系的隐私保护方法.现代计算机(专业版).2013,全文.* |
| Publication number | Publication date |
|---|---|
| CN112163160A (en) | 2021-01-01 |
| Publication | Publication Date | Title |
|---|---|---|
| CN112163160B (en) | Sensitive identification method based on knowledge graph | |
| US20240028651A1 (en) | System and method for processing documents | |
| US12248862B1 (en) | System for deep learning using knowledge graphs | |
| US12147433B2 (en) | Semantic entity search using vector space | |
| US20210158187A1 (en) | System and method for detecting friction in websites | |
| US20190384745A1 (en) | Systems and Methods for Management of Data Platforms | |
| US8589429B1 (en) | System and method for providing query recommendations based on search activity of a user base | |
| CN109145121B (en) | A fast storage and query method for time-varying graph data | |
| CN109446341A (en) | The construction method and device of knowledge mapping | |
| US20150356123A1 (en) | Systems and methods for management of data platforms | |
| CN117172725A (en) | Knowledge-graph-based industrial chain multi-cooperation intelligent decision method | |
| US11360953B2 (en) | Techniques for database entries de-duplication | |
| US8972463B2 (en) | Method and apparatus for functional integration of metadata | |
| US12086820B2 (en) | Technology opportunity mapping | |
| CN112560418B (en) | Create line item information from free-form tabular data | |
| CN106844407A (en) | Label network production method and system based on data set correlation | |
| US12072918B1 (en) | Machine learning using knowledge graphs | |
| CN106096609A (en) | A kind of merchandise query keyword automatic generation method based on OCR | |
| US11615089B1 (en) | System, method, and computer program for converting a natural language query to a structured database query | |
| US20210125294A1 (en) | Systems and methods for identifying potential real estate buyers | |
| US20240220876A1 (en) | Artificial intelligence (ai) based data product provisioning | |
| CN113190651B (en) | Electric power data global knowledge graph completion method based on quota knowledge graph technology | |
| EP3152678B1 (en) | Systems and methods for management of data platforms | |
| CN117726389A (en) | Cross-border e-commerce product purchase guiding system based on clustering recommendation algorithm | |
| US12340381B2 (en) | Linking physical locations and online channels in a database |
| 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |