CN105681908A - Broadcast television system based on individual watching behaviour and personalized programme recommendation method thereof - Google Patents

Abstract

Translated fromChinese

本发明公开了一种基于个体收视行为的广播电视系统及其个性节目推荐方法。本发明提供一种广播电视系统，包括：输入部，用于输入上述广播电视系统进行个性节目推荐所需的各种参数和各种指令；节目信息存储部，用于存储关于各种广播电视节目的信息及数据；分析单元，利用通过输入部输入的各种参数和从上述节目信息存储部读取的关于广播电视节目的信息，生成要发送的个性节目单，并且确定作为发送对象的推荐人群；以及推荐信息发送部，向确定的上述推荐人群发送上述个性节目单。根据该结构，能够向广播电视系统的不同用户，推荐不同的个性节目单。

The invention discloses a radio and television system based on individual viewing behavior and a method for recommending personalized programs. The present invention provides a radio and television system, including: an input unit for inputting various parameters and instructions required by the above-mentioned radio and television system for recommending personalized programs; a program information storage unit for storing information about various radio and television programs information and data; the analysis unit uses various parameters input through the input unit and the information about broadcast and television programs read from the program information storage unit to generate a personalized program list to be sent, and determine the recommended group of people as the sending target ; and the recommendation information sending unit, sending the above-mentioned personalized program list to the above-mentioned determined recommendation group. According to this configuration, it is possible to recommend different personalized program lists to different users of the broadcast television system.

Description

Translated fromChinese

基于个体收视行为的广播电视系统及其个性节目推荐方法Radio and television system and its personalized program recommendation method based on individual viewing behavior

本申请是于2013年7月1日递交的，申请号为201310272576.9，发明创造名称“广播电视系统及该系统中的个性节目推荐方法”的分案申请，本申请要求申请号为201210342540.9，申请日为2012年9月17日的中国专利申请的优先权。This application was submitted on July 1, 2013, the application number is 201310272576.9, the divisional application of the invention title "Broadcast TV system and the method for recommending personalized programs in the system", the application number is 201210342540.9, and the application date is Priority for the Chinese patent application dated September 17, 2012.

技术领域technical field

本发明涉及在广播电视技术领域，更详细的说，涉及能够向特定的收视群体分别推荐对应的个性节目的广播电视系统以及该广播电视系统中的个性节目推荐方法。The present invention relates to the technical field of broadcasting and television, and more specifically, relates to a broadcasting and television system capable of respectively recommending corresponding personalized programs to specific audience groups and a method for recommending personalized programs in the broadcasting and television system.

背景技术Background technique

随着电视节目日益丰富，广播电视用户正面临着与互联网用户类似的“信息过载”的问题，在这样的环境下，如何能够跟踪用户的兴趣变化，寻找用户感兴趣的电视节目内容的问题十分紧迫，广播电视个性节目推荐系统能够有效解决这一问题。With the increasing abundance of TV programs, radio and TV users are facing the same "information overload" problem as Internet users. In such an environment, how to track changes in users' interests and find the content of TV programs that users are interested in Urgent, radio and television personalized program recommendation system can effectively solve this problem.

广播电视个性节目推荐的理论基础为决策支持技术和数据挖掘技术。决策支持系统(DSS，DecisionSupportSystem)是由美国科学家Keen和ScottMorton于20世纪70年代首次提出的，到了20世纪80年代已经取得了巨大的发展。随着国内外专家学者的不断研究与探索，如今决策支持系统已经发展为数据仓库、联机分析处理和数据挖掘相结合的新型决策支持系统。它的典型特点是从海量数据中获取辅助决策的信息。数据挖掘(DM，DataMining)是从大量数据中提取有价值的知识的一门技术。随着数据挖掘技术的不断完善，数据挖掘在决策支持领域得到了越来越广泛的应用。这些知识为决策提供了有力的支持。广播电视个性节目推荐以决策支持系统为基础，构建解决问题的模型和方法，并通过数据挖掘技术挖掘用户收视行为规律和挖掘潜在收视人群。The theoretical basis of radio and television personalized program recommendation is decision support technology and data mining technology. Decision Support System (DSS, DecisionSupportSystem) was first proposed by American scientists Keen and ScottMorton in the 1970s, and has achieved great development in the 1980s. With the continuous research and exploration of experts and scholars at home and abroad, today's decision support system has developed into a new type of decision support system that combines data warehouse, online analysis processing and data mining. Its typical feature is to obtain auxiliary decision-making information from massive data. Data Mining (DM, DataMining) is a technology to extract valuable knowledge from a large amount of data. With the continuous improvement of data mining technology, data mining has been more and more widely used in the field of decision support. This knowledge provides strong support for decision-making. Radio and television personalized program recommendation is based on the decision support system, constructs models and methods to solve problems, and uses data mining technology to mine the rules of user viewing behavior and tap potential audiences.

个性节目推荐的本质是对用户收看的节目进行排序，在这一领域中，目前已有的方法有简单统计算法、简单级联聚类算法、Bayes网络算法、多重特征下的排序算法等。以上几种方法存在的共同问题是仅能够实现对用户收看节目的排序，但是不能针对不用特征的用户提供不同服务，同时不具备对收视用户分群的能力。The essence of personalized program recommendation is to sort the programs that users watch. In this field, the existing methods include simple statistical algorithm, simple cascade clustering algorithm, Bayesian network algorithm, sorting algorithm under multiple features, etc. The common problem in the above several methods is that they can only sort the programs that users watch, but they cannot provide different services for users with different characteristics, and they do not have the ability to group viewing users.

发明内容Contents of the invention

本发明为解决现有技术中的上述问题点而作出，其目的在于提供一种广播电视系统以及该广播电视系统中的个性节目推荐方法，可以根据收视用户的不同需求灵活地推荐广播电视节目，实现个性节目推荐的功能。The present invention is made to solve the above-mentioned problems in the prior art, and its purpose is to provide a radio and television system and a method for recommending personalized programs in the radio and television system, which can flexibly recommend radio and television programs according to the different needs of viewing users, Realize the function of personalized program recommendation.

为此，本发明提供一种广播电视系统，其包括：输入部，用于输入上述广播电视系统进行个性节目推荐所需的各种参数和各种指令；节目信息存储部，用于存储关于各种广播电视节目的信息及数据；分析单元，利用通过输入部输入的各种参数和从上述节目信息存储部读取的关于广播电视节目的信息，生成要发送的个性节目单，并且确定作为发送对象的推荐人群；以及推荐信息发送部，向确定的上述收视群体发送上述个性节目单。To this end, the present invention provides a radio and television system, which includes: an input unit for inputting various parameters and instructions required for the above-mentioned radio and television system to recommend personalized programs; a program information storage unit for storing information about each The information and data of various radio and television programs; the analysis unit uses various parameters input through the input unit and the information about the radio and television programs read from the above-mentioned program information storage unit to generate a personalized program list to be transmitted, and determine as the transmission target recommendation groups; and the recommendation information sending unit, which sends the above-mentioned personalized program list to the above-mentioned determined audience groups.

此外，本发明还提供一种广播电视系统中的个性节目推荐方法，该广播电视系统包括输入部、节目信息存储部、分析单元和推荐信息发送部，其特征在于，该方法包括以下步骤：通过输入部输入上述广播电视系统进行个性节目推荐所需的各种参数和各种指令；分析步骤，利用通过输入部输入的各种参数和从上述节目信息存储部读取的关于广播电视节目的信息，生成要发送的个性节目单，并且确定作为发送对象的推荐人群；以及通过推荐信息发送部，向确定的上述收视群体发送上述个性节目单。In addition, the present invention also provides a method for recommending personalized programs in a radio and television system, the radio and television system includes an input unit, a program information storage unit, an analysis unit and a recommendation information sending unit, and it is characterized in that the method includes the following steps: The input unit inputs various parameters and various instructions required for the above-mentioned radio and television system to recommend personalized programs; the analysis step uses various parameters input through the input unit and information about radio and television programs read from the above-mentioned program information storage unit , generating a personalized program list to be sent, and determining a recommended group of people to be sent; and sending the above-mentioned personalized program list to the determined above-mentioned audience group through the recommendation information sending unit.

有益效果：Beneficial effect:

本发明实现了根据广播电视用户不同需求灵活选择个性节目推荐方法的解决方案。提供的节目类型阈值分析方法或聚类分析方法，能够实现协助节目制作商稳定节目忠实观众，寻找节目潜在观众的目的。用户收视行为分析方法通过具体用户收视行为的分析，能够实现有效把握用户的收视偏好，推荐个性节目的目的。The invention realizes the solution of flexibly selecting individualized program recommendation methods according to the different needs of radio and television users. The provided program type threshold analysis method or cluster analysis method can realize the purpose of assisting program producers to stabilize loyal program viewers and find potential program viewers. The user viewing behavior analysis method can effectively grasp the user's viewing preference and recommend personalized programs through the analysis of specific user viewing behavior.

附图说明Description of drawings

图1是表示本发明涉及的广播电视系统100的具体结构的示意图。FIG. 1 is a schematic diagram showing a specific configuration of a broadcast television system 100 according to the present invention.

图2是表示上述广播电视系统100所执行的个性节目推荐过程的流程图。FIG. 2 is a flow chart showing the personalized program recommendation process performed by the broadcast television system 100 described above.

图3是表示上述广播电视系统100中的分析单元130执行的分析过程的第一个例子的流程图。FIG. 3 is a flowchart showing a first example of the analysis process performed by the analysis unit 130 in the broadcast television system 100 described above.

图4是表示上述广播电视系统100中的分析单元130执行的分析过程的第二个例子的流程图。FIG. 4 is a flowchart showing a second example of the analysis process performed by the analysis unit 130 in the broadcast television system 100 described above.

具体实施方式detailed description

下面，参考附图来描述本发明涉及的广播电视系统和该系统中的个性节目推荐方法的具体实施例。In the following, specific embodiments of the broadcast television system and the method for recommending personalized programs in the system involved in the present invention will be described with reference to the accompanying drawings.

图1是表示本发明涉及的广播电视系统100的具体结构的示意图。FIG. 1 is a schematic diagram showing a specific configuration of a broadcast television system 100 according to the present invention.

如图1所示，本发明的广播电视系统100包括输入部110、节目信息存储部120、分析单元130、推荐信息发送部140。As shown in FIG. 1 , the broadcast television system 100 of the present invention includes an input unit 110 , a program information storage unit 120 , an analysis unit 130 , and a recommendation information sending unit 140 .

其中，输入部110用于输入上述广播电视系统进行个性节目推荐所需的各种参数等数据和各种指令，其可以是键盘、触摸屏、手写输入设备、鼠标等。Wherein, the input unit 110 is used to input data such as various parameters and various instructions required by the above-mentioned radio and television system for recommending personalized programs, and it may be a keyboard, a touch screen, a handwriting input device, a mouse, and the like.

节目信息存储部120用于存储关于各种广播电视节目的信息，例如节目的类型、节目的时间参数、预先设定的各种阈值等。此外，上述节目信息存储部120还可以存储上述广播电视系统100执行功能所需的其他数据。这些信息和数据可以预先存储在上述节目信息存储部120，也可以由输入部110输入后存储在上述节目信息存储部120。The program information storage unit 120 is used to store information about various broadcast TV programs, such as program types, program time parameters, various preset thresholds, and the like. In addition, the program information storage unit 120 may also store other data required by the broadcast television system 100 to perform functions. These information and data may be stored in the program information storage unit 120 in advance, or may be stored in the program information storage unit 120 after being input by the input unit 110 .

分析单元130用于利用通过输入部110输入的各种参数等数据和从上述节目信息存储部120读取的关于广播电视节目的信息，对上述各种参数和信息进行分析处理，生成要发送的个性节目单，并且确定作为发送对象的推荐人群。在本发明中，分析单元130可以利用节目类型分析法(例如节目类型阈值分析法、节目类型聚类分析法)、收视行为分析法(例如收视个体行为分析法、收视群体行为分析法)中的任一种，其中，当利用节目阈值类型分析法或节目类型聚类分析法的情况下，可以通过计算出收视人群占有率来确定推荐人群，当利用收视个体行为分析法或收视群体行为分析法的情况下，可以通过计算出频道贡献率来确定推荐节目单的频道，从而生成用于推荐的个性节目单。上述分析单元130选择使用哪个分析法，可以由来自输入部110的用户指令确定。而且，关于这些分析法的具体内容将在下面详细论述。The analysis unit 130 is used to analyze and process the above-mentioned various parameters and information by using the data such as various parameters input through the input unit 110 and the information about the broadcast TV program read from the above-mentioned program information storage unit 120, and generate the data to be transmitted. Personalized program list, and determine the recommended group as the sending target. In the present invention, the analysis unit 130 can use program type analysis method (such as program type threshold analysis method, program type cluster analysis method), viewing behavior analysis method (such as viewing individual behavior analysis method, viewing group behavior analysis method) Either one, wherein, when using the program threshold type analysis method or the program type cluster analysis method, the recommended audience can be determined by calculating the share of the audience, when using the individual audience behavior analysis method or the audience group behavior analysis method In the case of , the channel of the recommended program list can be determined by calculating the channel contribution rate, so as to generate a personalized program list for recommendation. The analysis method selected by the analysis unit 130 may be determined by a user instruction from the input unit 110 . Moreover, the specific content of these analysis methods will be discussed in detail below.

推荐信息发送部140，通过短信方式或电子邮件等方式向确定的收视个体或群体发送上述个性节目单。在此，该推荐信息发送部140可以是短信发送平台，也可以是电子邮件发送平台。The recommendation information sending unit 140 sends the above-mentioned personalized program list to the determined viewing individual or group by means of short message or email. Here, the recommendation information sending unit 140 may be a short message sending platform, or an email sending platform.

接着，参照图2说明上述广播电视系统所100执行的个性节目推荐过程。Next, the personalized program recommendation process performed by the broadcast television system 100 will be described with reference to FIG. 2 .

首先，用户通过输入部110输入进行个性节目推荐所需的各种参数等数据和各种指令(步骤S211)。在此，这些参数可以包括用户所在地区的参数、节目的类型参数和时间参数等。各种指令可以包括用于选择分析单元130利用的分析方法的指令，还可以包括确定推荐信息发送部140的发送方式的指令。上述参数等数据可以存储在节目信息存储部120，也可以发送给分析单元130使用。First, the user inputs data such as various parameters and various instructions required for personalized program recommendation through the input unit 110 (step S211). Here, these parameters may include parameters of the region where the user is located, program type parameters and time parameters, and so on. The various instructions may include an instruction for selecting an analysis method used by the analysis unit 130 , and may also include an instruction for determining a transmission method of the recommended information transmission unit 140 . Data such as the above parameters can be stored in the program information storage unit 120, or can be sent to the analysis unit 130 for use.

接着，在步骤S212,分析单元130对上述各种参数和信息进行分析处理，生成要发送的个性节目单，并且确定作为发送对象的推荐个体或群体。在此，分析单元130可以根据来自输入部110的指令来确定使用哪种分析法。分析单元130确定作为推荐对象的收视个体或群体，并且生成向上述推荐个体或群体推荐的个性节目单。Next, in step S212, the analysis unit 130 analyzes and processes the above-mentioned various parameters and information, generates a personalized program list to be sent, and determines recommended individuals or groups to be sent. Here, the analysis unit 130 may determine which analysis method to use according to an instruction from the input unit 110 . The analysis unit 130 determines the individual or group of viewers to be recommended, and generates a personalized program list recommended to the above-mentioned recommended individual or group.

然后，在步骤S213，上述推荐信息发送部140按照所选择的的发送方式，向已确定的收视个体或群体发送个性节目单。Then, in step S213, the recommendation information sending unit 140 sends the personalized program list to the determined viewing individual or group according to the selected sending method.

需要说明的是，本说明书中记载的术语“人群”，可以是单独的收视个体，也可以是由多个收视个体组成的收视群体。It should be noted that the term "group of people" recorded in this specification may be a single viewing individual, or a viewing group composed of multiple viewing individuals.

下面，参照图3和图4详细说明上述分析单元130所执行的分析动作。图3是表示上述广播电视系统100中的分析单元130执行的分析过程的第一个例子的流程图。图4是表示上述广播电视系统100中的分析单元130执行的分析过程的第二个例子的流程图。Next, the analysis operation performed by the above analysis unit 130 will be described in detail with reference to FIG. 3 and FIG. 4 . FIG. 3 is a flowchart showing a first example of the analysis process performed by the analysis unit 130 in the broadcast television system 100 described above. FIG. 4 is a flowchart showing a second example of the analysis process performed by the analysis unit 130 in the broadcast television system 100 described above.

(第一例：节目类型分析法)(First example: program type analysis method)

首先，在步骤S311，进行参数选择。在此，上述参数包括地区参数、节目类型参数和时间参数。First, in step S311, parameter selection is performed. Here, the above parameters include region parameters, program type parameters and time parameters.

在本实施例中，地区参数可以选择数据库中存在的任何地区的数据，其中，上述数据是以省或市级数据为单位。In this embodiment, the region parameter can select the data of any region existing in the database, wherein the above data is in the unit of provincial or municipal data.

在本实施例中，节目类型参数可以可以选择两级节目分类或三级节目分类，其中节目类型的两级分类包括4类，分别为新闻类节目、娱乐类节目、教育类节目和服务类节目；上述两级节目分类的上述三级节目分类总共包括27类，其中上述新闻类节目的上述三级节目分类为综合新闻消息节目、分类新闻消息节目、新闻专题类节目、新闻谈话节目、国际新闻类节目、大型新闻节目；上述娱乐类节目的三级节目分类为电视剧节目、体育节目、电影类节目、综艺节目、音乐节目、戏剧节目、游戏节目、真人秀节目、娱乐谈话.专题节目、国际娱乐类节目、大型娱乐节目；上述教育类节目的上述三级节目分类为社会教育节目、少儿.青年节目、国际教育类节目、大型教育节目；上述服务类节目的三级节目分类为生活服务节目、理财节目、广告类节目、国家服务类节目、频道宣传.收视服务节目、大型服务节目。In this embodiment, the program type parameter can select two-level program classification or three-level program classification, wherein the two-level classification of program type includes 4 categories, namely news programs, entertainment programs, educational programs and service programs ; The above-mentioned three-level program classification of the above-mentioned two-level program classification includes a total of 27 categories, of which the above-mentioned three-level program classification of the above-mentioned news programs includes comprehensive news programs, classified news programs, special news programs, news talk programs, and international news programs. Classified programs, large-scale news programs; the third-level programs of the above-mentioned entertainment programs are classified into TV drama programs, sports programs, movie programs, variety shows, music programs, drama programs, game shows, reality shows, entertainment talks, special programs, international Entertainment programs and large-scale entertainment programs; the above-mentioned third-level programs of the above-mentioned educational programs are classified as social education programs, children and youth programs, international educational programs, and large-scale educational programs; the above-mentioned third-level programs of service programs are classified as life service programs , financial programs, advertising programs, national service programs, channel promotion, viewing service programs, and large-scale service programs.

另外，在本实施例中，上述时间参数所选择的分析时间段为一周，推荐节目单的时间段为当前时间的下一周。In addition, in this embodiment, the analysis time period selected by the time parameter is one week, and the time period of the recommended program list is the next week of the current time.

在步骤S312，根据上述节目类型生成个性节目单，并将该个性节目单作为当前时间的下一周的上述节目类型的节目单。In step S312, a personalized program list is generated according to the above-mentioned program type, and the personalized program list is used as the program list of the above-mentioned program type for the next week at the current time.

在步骤S313，根据上述地区参数、节目类型参数和时间参数来计算收视时长。上述收视时长可以通过下式取得：In step S313, the viewing duration is calculated according to the above-mentioned region parameter, program type parameter and time parameter. The above viewing duration can be obtained by the following formula:

$\mathrm{<span><span>T</span>T</span>}<span><span>=</span>=</span>\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}}$

其中：in:

n表示所选节目类型中的节目数量；n represents the number of programs in the selected program type;

T_i表示有效收视时长。T_i represents the effective viewing time.

在本实施例中，有效收视时长包括以下四种情况：In this embodiment, the effective viewing duration includes the following four situations:

当WR_Begin<TV_Begin,WR_End<TV_End时,有效收视时长定义为WR_End-TV_Begin；When WR_Begin<TV_Begin, WR_End<TV_End, the effective viewing duration is defined as WR_End-TV_Begin;

当TV_Begin<WR_Begin,TV_End<WR_End时,有效收视时长定义为TV_End-WR_Begin；When TV_Begin<WR_Begin, TV_End<WR_End, the effective viewing duration is defined as TV_End-WR_Begin;

当WR_Begin<TV_Begin,TV_End<WR_End时,有效收视时长定义为TV_End-TV_Begin；When WR_Begin<TV_Begin, TV_End<WR_End, the effective viewing duration is defined as TV_End-TV_Begin;

当TV_Begin<WR_Begin,WR_End<TV_End时,有效收视时长定义为WR_End-WR_Begin。When TV_Begin<WR_Begin, WR_End<TV_End, the effective viewing duration is defined as WR_End-WR_Begin.

其中，WR_Begin和WR_End分别表示满足条件的收视纪录开始和终止的时间(WR_Begin<WR_End)；Among them, WR_Begin and WR_End represent the time of the beginning and end of the viewing record that meets the conditions respectively (WR_Begin<WR_End);

TV_Begin和TV_End分别表示某节目播出开始和终止的时间(TV_Begin<TV_End)；TV_Begin and TV_End indicate the start and end time of a certain program respectively (TV_Begin<TV_End);

步骤S314，根据计算出的上述收视时长进行收视人群分析。Step S314, analyzing audience groups according to the calculated viewing duration.

本实施例中，分析上述收视人群可以使用节目类型阈值分析法或节目类型聚类分析法中的任一种。In this embodiment, any one of the program type threshold analysis method or the program type cluster analysis method may be used to analyze the above audience group.

上述节目类型阈值分析法包括以下步骤：The above program type threshold analysis method comprises the following steps:

利用预先设定的两个阈值i和j(0<i<j<最大收视时长)，将收视人群进行以下分类：Using two pre-set thresholds i and j (0<i<j<maximum viewing duration), the viewing audience is classified as follows:

当T≤i时，为流失家庭，即很少观看上述节目类型的潜在家庭；When T≤i, it is a lost family, that is, a potential family who seldom watches the above program types;

当i<T<j时，为普通家庭，即对上述类型节目关注较多，又不是十分热衷的普通家庭；When i<T<j, it is an ordinary family, that is, an ordinary family that pays more attention to the above-mentioned types of programs but is not very enthusiastic;

当j≤T时，为忠实家庭，即对上述类型节目十分关注的忠实家庭；其中，T为上述收视时长。When j≤T, it is a loyal family, that is, a loyal family that pays great attention to the above-mentioned types of programs; where T is the above-mentioned viewing time.

此外，上述节目类型聚类分析法包括以下步骤：设定待聚类的上述收视时长个数为n(n>0)，上述收视时长的人群聚类个数k(0<k≤n)，In addition, the above-mentioned program type clustering analysis method includes the following steps: setting the number of the above-mentioned viewing durations to be clustered as n (n>0), the number of clusters of the above-mentioned viewing durations k (0<k≤n),

步骤A：随机选取k个上述收视时长作为各簇的初始均值，并将上述均值定义为各簇上述收视时长的统计平均值，Step A: Randomly select k above-mentioned viewing durations as the initial average value of each cluster, and define the above-mentioned average value as the statistical average value of the above-mentioned viewing durations of each cluster,

步骤B：定义均方误差EStep B: Define the mean square error E

$\mathrm{<span><span>E</span>E.</span>}<span><span>=</span>=</span>\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>k</span>k</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>T</span>T</span>}<span><span>\&Element;</span>\&Element;</span>\mathrm{<span><span>C</span>C</span>}\mathrm{<span><span>i</span>i</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}<span><span>|</span>|</span>\mathrm{<span><span>T</span>T</span>}<span><span>-</span>-</span>{\mathrm{<span><span>m</span>m</span>}}_{\mathrm{<span><span>i</span>i</span>}}{<span><span>|</span>|</span>}^{\mathrm{<span><span>2</span>2</span>}}$

其中：in:

T表示上述收视时长；T represents the above viewing time;

C_i表示某个簇；C_i represents a cluster;

m_i表示簇C_i的上述收视时长的均值；m_i represents the mean value of the above-mentioned viewing time of cluster C_i ;

根据如上定义，计算每个簇中的各个上述收视时长的上述均方误差E，并将每个上述收视时长指派到最相似的簇，即与簇均值的距离最小的簇；According to the above definition, calculate the above-mentioned mean square error E of each of the above-mentioned viewing durations in each cluster, and assign each of the above-mentioned viewing durations to the most similar cluster, that is, the cluster with the smallest distance from the cluster mean value;

步骤C：对于更新后的簇，计算每个簇中上述收视时长的均值；Step C: For the updated clusters, calculate the mean value of the above-mentioned viewing time in each cluster;

步骤D：重复步骤B至步骤C，直至更新后的簇不再发生变化，则得到经过上述聚类分析处理后的k个簇，即k类人群；Step D: Repeat Step B to Step C until the updated clusters no longer change, then k clusters after the above cluster analysis processing are obtained, that is, k groups of people;

按照各簇均值由大到小进行排序，依次定义各簇为级别1家庭、级别2家庭……其中级别1家庭对此类节目忠实度最高，此后级别忠实度依次递减。According to the mean values of each cluster, they are sorted from large to small, and each cluster is defined in turn as level 1 family, level 2 family... Among them, level 1 family has the highest loyalty to this type of program, and then the level of loyalty decreases in descending order.

接着，在步骤S315，根据上述阈值分析得出的上述收视人群来计算收视人群占有率。上述收视人群占有率可以利用下式取得：Next, in step S315, the audience share is calculated according to the above-mentioned audience audience obtained from the above-mentioned threshold analysis. The audience share mentioned above can be obtained using the following formula:

$\mathrm{<span><span>P</span>P</span>}\mathrm{<span><span>E</span>E.</span>}\mathrm{<span><span>R</span>R</span>}<span><span>=</span>=</span>\frac{{\mathrm{<span><span>N</span>N</span>}}^{<span><span>\&prime;</span>\&prime;</span>}}{\mathrm{<span><span>N</span>N</span>}}<span><span>\&times;</span>\&times;</span>\mathrm{<span><span>100</span>100</span>}\mathrm{<span><span>\%</span>\%</span>}$

其中：in:

N'表示某类人群的收视户数(0≤N')；N' indicates the number of viewers of a certain group of people (0≤N');

N表示所选地区的收视总户数(N'≤N)。N represents the total number of viewers in the selected area (N'≤N).

在步骤S316，参照上述收视人群占有率，确定作为发送个性节目单的推荐对象的推荐人群。其中，上述推荐人群可以是任意人群或任意几个人群的组合。这样，就由分析单元130生成了要发送的个性节目单，并且确定了作为发送对象的推荐人群。In step S316, referring to the above-mentioned viewer group share, determine the recommended group as the recommendation object for sending the personalized program list. Wherein, the above-mentioned recommended group of people may be any group of people or a combination of any number of groups of people. In this way, the analysis unit 130 generates a personalized program list to be sent, and determines the recommended group of people to be sent.

(第二例：收视行为分析法)(Second example: viewing behavior analysis method)

收视行为分析法包括个体收视行为分析法和群体收视行为分析法，其中，个体收视行为分析法是选择某一个体的收视信息进行分析计算，取得向该个体推荐节目单的推荐频道；群体收视行为分析法是选择某一收视群体的收视信息进行分析计算，取得向该群体推荐节目单的推荐频道。下面，参照图4分别说明上述个体收视行为分析法和群体收视行为分析法的具体分析步骤。Viewing behavior analysis method includes individual viewing behavior analysis method and group viewing behavior analysis method, wherein, the individual viewing behavior analysis method is to select a certain individual’s viewing information for analysis and calculation, and obtain the recommended channel for recommending the program list to the individual; group viewing behavior The analysis method is to select the rating information of a certain audience group for analysis and calculation, and obtain the recommended channels for recommending program lists to this group. Next, with reference to FIG. 4 , the specific analysis steps of the individual viewing behavior analysis method and the group viewing behavior analysis method are described respectively.

1、个体收视行为分析法1. Individual viewing behavior analysis method

下面，参照图4说明个体收视行为分析法的具体过程。Next, the specific procedure of the individual viewing behavior analysis method will be described with reference to FIG. 4 .

首先，在步骤S411，选择作为分析对象的具体用户和时间参数，并提取上述具体用户的信息。First, in step S411, select specific users and time parameters as analysis objects, and extract information about the above specific users.

接着，在步骤S412，计算上述具体用户的收视时长，上述收视时长可以通过下式取得：Next, in step S412, the viewing duration of the above-mentioned specific user is calculated, and the above-mentioned viewing duration can be obtained by the following formula:

${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>1</span>1</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>l</span>l</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>4</span>4</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>2</span>2</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>l</span>l</span>}}$

${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>2</span>2</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>2</span>2</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}}$

其中：in:

T_s1表示上述具体用户收看所有节目的收视总时长；T_s1 represents the total viewing time of all programs watched by the above-mentioned specific users;

T_s2表示上述具体用户收看某类节目的收视总时长；T_s2 represents the total viewing time of the specific user watching a certain type of program;

n₁表示某日上述具体用户收看某频道某类节目的节目数目；n₁ represents the number of programs of a certain type of program on a certain channel watched by the above-mentioned specific user on a certain day;

n₂表示某日上述具体用户收看某类节目的频道数目；n₂ represents the number of channels that the above-mentioned specific user watches a certain type of program on a certain day;

n₃表示分析具体用户上述第三收视时长的输入天数；n₃ represents the number of input days for analyzing the above-mentioned third viewing duration of a specific user;

n₄表示上述具体用户收看节目类型总数；n₄ represents the total number of program types watched by the above-mentioned specific users;

T_i,j,k,l和T_i,j,k表示收看某具体节目的有效收视时长。T_{i, j, k, l} and T_{i, j, k} represent the effective viewing time for watching a specific program.

然后，在步骤S413，利用下式计算该具体用户的收视偏好：Then, in step S413, utilize following formula to calculate the viewing preference of this specific user:

并将上述收视偏好按由大到小排序，对上述具体用户推荐收视偏好最大的节目类型。And the above-mentioned viewing preference is sorted from large to small, and the program type with the largest viewing preference is recommended to the above-mentioned specific user.

在步骤S414，根据上述收视偏好的计算结果，选取上述收视偏好最大的节目类型，利用计算上述收视时长得到的结果，利用下式计算频道贡献率：In step S414, according to the calculation result of the above-mentioned viewing preference, select the program type with the largest viewing preference, and use the result obtained by calculating the above-mentioned viewing duration to calculate the channel contribution rate using the following formula:

${\mathrm{<span><span>\&psi;</span>\&psi;</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}}<span><span>=</span>=</span>\frac{{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>3</span>3</span>}}}{{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>2</span>2</span>}}}$

其中：in:

${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>3</span>3</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}}$

T_s3表示上述用具体户收看某频道该类节目的收视时长；T_s3 represents the viewing time of the above-mentioned specific user watching this type of program on a certain channel;

n₁表示某日上述具体用户收看某频道该类节目的节目数目；n₁ represents the number of programs of this type of programs on a certain channel that the above-mentioned specific users watch on a certain day;

n₃表示分析上述具体用户的收视时长的输入天数；n₃ represents the number of input days for analyzing the viewing duration of the above-mentioned specific users;

T_i,k表示收看某具体节目的有效收视时长。T_i,k represents the effective viewing time for watching a specific program.

然后，在步骤S415，将上述频道贡献率进行排序，根据上述频道贡献率和上述收视偏好生成推荐节目单，确定推荐上述推荐节目单的频道，其中，上述频道可以选择任意频道贡献率的单个频道或频道组合。Then, in step S415, the above-mentioned channel contribution rates are sorted, a recommended program list is generated according to the above-mentioned channel contribution rates and the above-mentioned viewing preferences, and the channels for recommending the above-mentioned recommended program list are determined, wherein the above-mentioned channel can select a single channel with any channel contribution rate or channel combinations.

2、群体收视行为分析法2. Group viewing behavior analysis method

群体收视行为分析法和个体收视行为分析法的大体步骤相似，因此，下面同样参照图4说明群体收视行为分析法的具体过程。The general steps of the group viewing behavior analysis method and the individual viewing behavior analysis method are similar. Therefore, the specific process of the group viewing behavior analysis method will also be described below with reference to FIG. 4 .

首先，在步骤S411，选择作为分析对象的收视群体类型和时间参数，并提取该收视群体的信息，例如，群体分类包含房地产/建筑、服务业、工业/地质、广播电视/文化艺术、计算机(IT/互联网)、交通/运输、教育/培训、金融(银行/证券/保险)、酒店/旅游/餐饮、贸易/进出口、媒体/广告/咨询、农业/水产、退休、学生、医疗/保健/制药、政府机关和其他等共17类人群。At first, in step S411, select audience group type and time parameter as analysis object, and extract the information of this audience group, for example, group classification includes real estate/building, service industry, industry/geology, radio/television/culture art, computer ( IT/Internet), transportation/transportation, education/training, finance (banking/securities/insurance), hotel/tourism/catering, trade/import, media/advertising/consulting, agriculture/aquaculture, retirement, students, medical care/health care /pharmaceuticals, government agencies and others, a total of 17 groups of people.

接着，在步骤S412，利用下式计算所选择的上述收视群体的收视时长：Next, in step S412, use the following formula to calculate the viewing duration of the selected above-mentioned audience groups:

${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>c</span>c</span>}\mathrm{<span><span>1</span>1</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>p</span>p</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>5</span>5</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>l</span>l</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>4</span>4</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>2</span>2</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>l</span>l</span>}<span><span>,</span>,</span>\mathrm{<span><span>p</span>p</span>}}$

${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>c</span>c</span>}\mathrm{<span><span>2</span>2</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>p</span>p</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>5</span>5</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>2</span>2</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>p</span>p</span>}}$

其中：in:

T_c1表示上述收视群体收看所有节目的收视总时长；T_c1 represents the total viewing time of all programs watched by the above audience groups;

T_c2表示上述收视群体收看某类节目的收视总时长；T_c2 represents the total viewing time of the above-mentioned audience group watching a certain type of program;

n₁表示某日上述收视群体收看某频道的某类节目的节目数目；n₁ represents the number of programs of a certain type of programs on a certain channel watched by the above audience group on a certain day;

n₂表示某日上述收视群体收看某类节目的频道数目；n₂ represents the number of channels of a certain program watched by the above audience group on a certain day;

n₃表示分析上述收视群体收视时长的输入天数；n₃ represents the number of input days for analyzing the viewing duration of the above audience groups;

n₄表示上述收视群体收看节目类型的总数；n₄ represents the total number of program types watched by the above audience groups;

n₅表示上述收视群体包含的收视个体人数；n₅ represents the number of individual viewers contained in the above-mentioned audience groups;

T_i,j,k,l,p和T_i,j,k,p表示收看某具体节目的有效收视时长。T_{i, j, k, l, p} and T_{i, j, k, p} represent the effective viewing time for watching a specific program.

接着，在步骤S413，利用下式计算上述收视群体的收视偏好Next, in step S413, use the following formula to calculate the viewing preferences of the above-mentioned viewing group

并且，将收视偏好按由大到小的顺序排序，向群体用户推荐收视偏好最大的节目类型。And, the viewing preference is sorted in descending order, and the program type with the largest viewing preference is recommended to group users.

然后，在步骤S414，根据计算出的上述收视偏好，选取收视偏好最大的节目类型，并且利用收视时长计算得到的结果，基于下式计算频道贡献率Ψ_c：Then, in step S414, according to the above-mentioned calculated viewing preference, select the program type with the largest viewing preference, and use the result obtained from the calculation of the viewing duration to calculate the channel contribution rate Ψ_c based on the following formula:

${\mathrm{<span><span>\&psi;</span>\&psi;</span>}}_{\mathrm{<span><span>c</span>c</span>}}<span><span>=</span>=</span>\frac{{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>c</span>c</span>}\mathrm{<span><span>3</span>3</span>}}}{{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>c</span>c</span>}\mathrm{<span><span>2</span>2</span>}}}$

其中：in:

${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>c</span>c</span>}\mathrm{<span><span>3</span>3</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>p</span>p</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>5</span>5</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>p</span>p</span>}}$

T_c3表示上述收视群体收看某频道该类节目的收视时长；T_c3 represents the viewing time of the above-mentioned audience group watching this type of program on a certain channel;

n₁表示某日上述收视群体收看该类节目的节目数目；n₁ represents the number of programs of this type of programs watched by the above-mentioned audience group on a certain day;

n₃表示分析上述收视群体收视时长的输入天数；n₃ represents the number of input days for analyzing the viewing duration of the above audience groups;

n₅表示上述收视群体中包含的个体人数；n₅ represents the number of individuals included in the above audience groups;

T_i,k,p表示收看某具体节目的有效收视时长。T_i,k,p represent the effective viewing time for watching a specific program.

最后，在步骤S415，将频道贡献率进行排序，所得值越大频道贡献率越高，而且，根据频道贡献率，选择推荐节目单的频道，其中，可以选择任意频道贡献率的单个频道或频道组合。Finally, in step S415, the channel contribution rate is sorted, the greater the value obtained, the higher the channel contribution rate, and, according to the channel contribution rate, select the channel for the recommended program list, where any single channel or channel with any channel contribution rate can be selected combination.

根据上述方法，可以根据广播电视用户不同需求灵活选择个性节目推荐方法，以达到推介个性节目的目的。According to the above method, a personalized program recommendation method can be flexibly selected according to different needs of radio and television users, so as to achieve the purpose of recommending personalized programs.

优选地，本实施例上述的用户可以是个体用户，也可以是群体用户。Preferably, the above-mentioned users in this embodiment may be individual users or group users.

根据如上上述的本发明，实现了根据广播电视用户不同需求灵活选择个性节目推荐方法的解决方案。提供的节目类型阈值分析方法或聚类分析方法，能够实现协助节目制作商稳定节目忠实观众，寻找节目潜在观众的目的。用户收视行为分析方法通过具体用户收视行为的分析，能够实现有效把握用户的收视偏好，推荐个性节目的目的。According to the above-mentioned present invention, a solution for flexibly selecting personalized program recommendation methods according to different needs of broadcasting and television users is realized. The provided program type threshold analysis method or cluster analysis method can realize the purpose of assisting program producers to stabilize loyal program viewers and find potential program viewers. The user viewing behavior analysis method can effectively grasp the user's viewing preference and recommend personalized programs through the analysis of specific user viewing behavior.

在本发明的上述教导下，本领域技术人员可以在上述实施例的基础上对本发明所提供的广播电视的个性节目推荐方法和系统进行改进，而这些改进都落在本发明的保护范围内。本领域技术人员应该明白，上述的具体描述只是更好地解释本发明的目的，本发明的保护范围由权利要求及其等同物限定。Under the above-mentioned teaching of the present invention, those skilled in the art can improve the method and system for recommending personalized broadcast television programs provided by the present invention on the basis of the above-mentioned embodiments, and these improvements all fall within the protection scope of the present invention. Those skilled in the art should understand that the above detailed description is only to better explain the purpose of the present invention, and the protection scope of the present invention is defined by the claims and their equivalents.

Claims (4)

Translated fromChinese

1.一种广播电视系统，其特征在于，包括：1. A broadcast television system, characterized in that, comprising:输入部，用于输入上述广播电视系统进行个性节目推荐所需的各种参数和各种指令；The input unit is used to input various parameters and various instructions required for the above-mentioned radio and television system to recommend personalized programs;节目信息存储部，用于存储关于各种广播电视节目的信息及数据；The program information storage unit is used to store information and data about various broadcast and TV programs;分析单元，利用通过输入部输入的各种参数和从上述节目信息存储部读取的关于广播电视节目的信息，生成要发送的个性节目单，并且确定作为发送对象的推荐人群，其中，分析单元利用个体收视行为分析法，通过频道贡献率来确定推荐节目单的频道，从而生成用于推荐的个性节目单，所述个体收视行为分析法是选择某一个体的收视信息进行分析计算，取得向该个体推荐节目单的推荐频道；以及The analysis unit uses various parameters input through the input unit and the information about the broadcast TV program read from the above-mentioned program information storage unit to generate a personalized program list to be sent, and determine the recommended group of people who are to be sent, wherein the analysis unit Using the individual viewing behavior analysis method, the channel of the recommended program list is determined through the channel contribution rate, thereby generating a personalized program list for recommendation. the recommended channel of the individual recommended program list; and推荐信息发送部，向确定的上述推荐人群发送上述个性节目单。The recommendation information sending unit sends the above-mentioned personalized program list to the above-mentioned determined recommendation group.2.如权利要求1所述的广播电视系统，其特征在于，上述分析单元执行以下分析过程：2. The broadcast television system as claimed in claim 1, wherein the analysis unit performs the following analysis process:(1)选择具体用户和时间参数，并提取上述具体用户的信息；(1) Select specific users and time parameters, and extract the information of the above specific users;(2)利用下式计算上述具体用户的收视时长：(2) Use the following formula to calculate the viewing time of the above-mentioned specific users:${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>1</span>1</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>l</span>l</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>4</span>4</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>2</span>2</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>l</span>l</span>}}$${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>2</span>2</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>2</span>2</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}}$其中：in:T_s1表示上述具体用户收看所有节目的收视总时长；T_s1 represents the total viewing time of all programs watched by the above-mentioned specific users;T_s2表示上述具体用户收看某类节目的收视总时长；T_s2 represents the total viewing time of the specific user watching a certain type of program;n₁表示某日上述具体用户收看某频道某类节目的节目数目；n₁ represents the number of programs of a certain type of program on a certain channel watched by the above-mentioned specific user on a certain day;n₂表示某日上述具体用户收看某类节目的频道数目；n₂ represents the number of channels that the above-mentioned specific user watches a certain type of program on a certain day;n₃表示分析上述具体用户上述收视时长的输入天数；n₃ represents the number of input days for analyzing the above-mentioned viewing duration of the above-mentioned specific users;n₄表示上述具体用户收看节目类型总数；n₄ represents the total number of program types watched by the above-mentioned specific users;T_i,j,k,l和T_i,j,k表示收看某具体节目的有效收视时长；T_{i, j, k, l} and T_{i, j, k} represent the effective viewing time for watching a specific program;(3)利用下式计算收视偏好：(3) Use the following formula to calculate viewing preference:并将上述收视偏好按由大到小排序；And sort the above viewing preferences in descending order;(4)根据上述收视偏好的计算结果，选取上述收视偏好最大的节目类型，利用计算上述收视时长得到的结果，按下式计算频道贡献率：(4) According to the calculation results of the above-mentioned viewing preference, select the program type with the largest viewing preference above, and use the result obtained by calculating the above-mentioned viewing duration to calculate the channel contribution rate according to the following formula:${\mathrm{<span><span>\&psi;</span>\&psi;</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}}<span><span>=</span>=</span>\frac{{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>3</span>3</span>}}}{{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>2</span>2</span>}}}$其中：in:${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>3</span>3</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}}$T_s3表示上述具体用户收看某频道该类节目的收视时长；T_s3 represents the viewing time of the above-mentioned specific user watching this type of program on a certain channel;n₁表示某日上述具体用户收看某频道该类节目的节目数目；n₁ represents the number of programs of this type of programs on a certain channel that the above-mentioned specific users watch on a certain day;n₃表示分析上述具体用户的收视时长的输入天数；n₃ represents the number of input days for analyzing the viewing duration of the above-mentioned specific users;T_i,k表示收看某具体节目的有效收视时长；T_i,k represents the effective viewing time for watching a specific program;(5)将上述频道贡献率进行排序，根据上述频道贡献率和上述收视偏好生成推荐节目单，并确定推荐上述推荐节目单的频道，其中，上述频道选择任意频道贡献率的单个频道或频道组合。(5) Sorting the above-mentioned channel contribution rates, generating a recommended program list according to the above-mentioned channel contribution rates and the above-mentioned viewing preference, and determining the channels for recommending the above-mentioned recommended program list, wherein, the above-mentioned channel selects a single channel or a combination of channels with any channel contribution rate .3.一种广播电视系统中的个性节目推荐方法，该广播电视系统包括输入部、节目信息存储部、分析单元和推荐信息发送部，其特征在于，该方法包括以下步骤：3. A method for recommending personalized programs in a radio and television system, the radio and television system comprising an input unit, a program information storage unit, an analysis unit and a recommended information sending unit, characterized in that the method comprises the following steps:通过输入部输入上述广播电视系统进行个性节目推荐所需的各种参数和各种指令；Input various parameters and various instructions required by the above-mentioned radio and television system for personalized program recommendation through the input unit;分析步骤，利用通过输入部输入的各种参数和从上述节目信息存储部读取的关于广播电视节目的信息，生成要发送的个性节目单，并且确定作为发送对象的推荐人群，其中，分析单元利用个体收视行为分析法，通过频道贡献率来确定推荐节目单的频道，从而生成用于推荐的个性节目单；以及通过推荐信息发送部，向确定的上述推荐人群发送上述个性节目单。The analysis step is to use various parameters input through the input unit and the information about the broadcast TV program read from the above-mentioned program information storage unit to generate a personalized program list to be sent, and to determine the recommended group of people as the sending object, wherein the analysis unit Using the individual viewing behavior analysis method to determine the channel of the recommended program list through the channel contribution rate, thereby generating a personalized program list for recommendation; and sending the above-mentioned personalized program list to the determined recommended group of people through the recommendation information sending unit.4.如权利要求3所述的个性节目推荐方法，其特征在于，上述分析步骤执行以下分析过程：4. The personalized program recommendation method as claimed in claim 3, characterized in that, said analysis step performs the following analysis process:(1)选择具体用户和时间参数，并提取上述具体用户的信息；(1) Select specific users and time parameters, and extract the information of the above specific users;(2)利用下式计算上述具体用户的收视时长：(2) Use the following formula to calculate the viewing time of the above-mentioned specific users:${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>1</span>1</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>l</span>l</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>4</span>4</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>2</span>2</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}<span><span>,</span>,</span>\mathrm{<span><span>l</span>l</span>}}$${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>2</span>2</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>j</span>j</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>2</span>2</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>j</span>j</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}}$其中：in:T_s1表示上述具体用户收看所有节目的收视总时长，T_s1 represents the total viewing time of all programs watched by the above-mentioned specific users,T_s2表示上述具体用户收看某类节目的收视总时长，T_s2 represents the total viewing time of the specific user watching a certain type of program,n₁表示某日上述具体用户收看某频道某类节目的节目数目，n₁ represents the number of programs of a certain type of program on a certain channel watched by the above-mentioned specific user on a certain day,n₂表示某日上述具体用户收看某类节目的频道数目，n₂ represents the number of channels of a certain type of program watched by the above-mentioned specific user on a certain day,n₃表示分析具体用户上述收视时长的输入天数，n₃ represents the number of input days for analyzing the above viewing duration of specific users,n₄表示上述具体用户收看节目类型总数，n₄ represents the total number of program types watched by the above-mentioned specific users,T_i,j,k,l和T_i,j,k表示收看某具体节目的有效收视时长，T_{i, j, k, l} and T_{i, j, k} represent the effective viewing time of a specific program,(3)利用下式计算收视偏好：(3) Use the following formula to calculate viewing preference:并将上述收视偏好按由大到小排序；And sort the above viewing preferences in descending order;(4)根据上述收视偏好的计算结果，选取上述收视偏好最大的节目类型，利用计算上述收视时长得到的结果，按下式计算频道贡献率：(4) According to the calculation result of the above-mentioned viewing preference, select the program type with the largest viewing preference above, and use the result obtained by calculating the above-mentioned viewing duration to calculate the channel contribution rate according to the following formula:${\mathrm{<span><span>\&psi;</span>\&psi;</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}}<span><span>=</span>=</span>\frac{{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>3</span>3</span>}}}{{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>2</span>2</span>}}}$其中：in:${\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>s</span>the\; s</span>}\mathrm{<span><span>3</span>3</span>}}<span><span>=</span>=</span>\underset{\mathrm{<span><span>k</span>k</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>3</span>3</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}\underset{\mathrm{<span><span>i</span>i</span>}<span><span>=</span>=</span>\mathrm{<span><span>1</span>1</span>}}{\overset{\mathrm{<span><span>n</span>no</span>}\mathrm{<span><span>1</span>1</span>}}{<span><span>\&Sigma;</span>\&Sigma;</span>}}{\mathrm{<span><span>T</span>T</span>}}_{\mathrm{<span><span>i</span>i</span>}<span><span>,</span>,</span>\mathrm{<span><span>k</span>k</span>}}$T_s3表示上述用户收看某频道该类节目的收视时长，T_s3 represents the viewing time of the above-mentioned user watching this type of program on a certain channel,n₁表示某日上述用户收看某频道该类节目的节目数目，n₁ represents the number of programs of this type of program on a certain channel that the above-mentioned user watched on a certain day,n₃表示分析上述用户的收视时长的输入天数，n₃ represents the number of input days for analyzing the viewing time of the above-mentioned users,T_i,k表示收看某具体节目的有效收视时长，T_i,k represents the effective viewing time for watching a specific program,(5)将上述频道贡献率进行排序，根据上述频道贡献率和上述收视偏好生成推荐节目单，并确定推荐上述推荐节目单的频道，其中，上述频道选择任意频道贡献率的单个频道或频道组合。(5) Sorting the above-mentioned channel contribution rates, generating a recommended program list according to the above-mentioned channel contribution rates and the above-mentioned viewing preference, and determining the channels for recommending the above-mentioned recommended program list, wherein, the above-mentioned channel selects a single channel or a combination of channels with any channel contribution rate .