Content accurate screening method and system based on big data behavior feature analysis

By analyzing big data behavioral characteristics, the content keyword index and user push coefficient are calculated, solving the problem of inaccurate filtering in existing technologies and achieving accurate filtering and optimized content dissemination.

CN118690150BActive Publication Date: 2024-11-22HANGZHOU JIUYANG CULTURE MEDIA CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410741978.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-06-11
Publication Date
2024-11-22
Estimated Expiration
2044-06-11

AI Technical Summary

Technical Problem

Existing technologies cannot accurately evaluate the content to be pushed based on the different times a user browses, cannot accurately determine the degree of matching between the user's historical search keywords and the user, and cannot perform precise filtering.

Method used

By using big data behavioral feature analysis, we obtain users' historical browsing data, calculate the first push index, second push index, and keyword user push coefficient for content keywords, and use the content push matching index for precise filtering.

Benefits of technology

It enables precise filtering of push content, improving user experience and optimizing content dissemination.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118690150B_ABST
    Figure CN118690150B_ABST
Patent Text Reader

Abstract

The present invention discloses a method and system for accurately screening content based on big data behavior feature analysis, and relates to the field of data push technology, including obtaining user historical browsing data, obtaining user historical browsing content keyword information based on the user historical browsing data, and obtaining a content push matching index based on the content information to be pushed, a content keyword first push index, a content keyword second push index, and a keyword user push coefficient. The present invention analyzes user browsing content preferences in different time periods through the content keyword first push index, analyzes user behavior characteristics when browsing content through the content keyword second push index, accurately evaluates user historical search keywords through the keyword user push coefficient, and accurately screens the content to be pushed through the content push matching index, thereby improving user experience and optimizing content dissemination effects.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of data push technology, specifically to a method and system for precise content filtering based on big data behavioral feature analysis. Background Technology

[0002] With the development of information technology and the Internet, people have gradually moved from an era of information scarcity to an era of information overload. At the same time, both information consumers and information producers have encountered great challenges: for information consumers, among the massive amount of multimedia content, users often find it difficult to find content that interests them and need to search manually. This not only reduces the user experience but also easily leads to users losing interest in browsing.

[0003] Currently, there are still some issues with accurate content filtering, such as the inability to accurately evaluate the content to be pushed based on different browsing times, the inability to accurately evaluate the keywords of the content based on the browsing methods and times of different users, the inability to accurately determine the matching degree between the user's historical search keywords and the user, and the inability to accurately filter the content to be pushed based on the keywords of the content to be pushed. Summary of the Invention

[0004] To address the aforementioned technical issues, this technical solution provides a method and system for precise content filtering based on big data behavioral feature analysis. It resolves the problems mentioned in the background section regarding the inability to accurately evaluate content to be pushed based on different user browsing times, the inability to accurately evaluate the keywords of browsed content based on different browsing methods and times, the inability to accurately determine the matching degree between the user's historical search keywords and the user, and the inability to accurately filter content to be pushed based on the keywords of the content to be pushed.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] Content filtering methods and systems based on big data behavioral feature analysis include:

[0007] Obtain user's historical browsing data, which includes user's historical browsing time information, user's historical browsing time information, and user's historical browsing content information;

[0008] Based on the user's historical browsing data, obtain the keyword information of the user's historical browsing content. The keyword information of the user's historical browsing content includes the keyword information of browsing content features and the keyword information of browsing content attributes.

[0009] Based on the user's historical browsing data, obtain the user's historical browsing time classification information;

[0010] Based on the user's historical browsing time category information and the user's historical browsing content keyword information, obtain the first push index of content keywords;

[0011] Based on users' historical browsing data, obtain user behavior characteristic information;

[0012] Based on user behavior characteristics and user browsing history keyword information, obtain the second push index for content keywords;

[0013] Based on the user's historical browsing data, the user's historical search information is obtained, which includes search keyword information and search time information;

[0014] Based on users' historical search information, obtain the keyword user push coefficient;

[0015] Obtain information about the content to be pushed, including keyword information for the content to be pushed;

[0016] Based on the information of the content to be pushed, the first push index of the content keywords, the second push index of the content keywords, and the keyword user push coefficient, obtain the content push matching index;

[0017] Based on the content matching index, the content to be pushed is accurately filtered.

[0018] Preferably, the step of obtaining the first push index of content keywords based on the user's historical browsing time classification information and the user's historical browsing content keyword information specifically includes:

[0019] Based on the user's browsing history data, obtain the user's browsing history time information;

[0020] Based on the user's historical browsing time information, obtain the user's basic browsing time period information;

[0021] Based on the keyword information of the user's historical browsing content, obtain the keyword information of the browsing content attributes;

[0022] Based on the user's basic browsing time period information and the keyword information of the browsing content attributes, the browsing content attribute change information is obtained. The browsing content attribute change information includes the proportion change information of the user's historical browsing content with different attributes within the basic browsing time period for each user.

[0023] Based on the need for precise content filtering, obtain the threshold for changes in the attributes of the browsed content;

[0024] Based on the information on changes in the attributes of the browsed content and the threshold for changes in the attributes of the browsed content, information on the user's browsing time period is obtained;

[0025] Based on the user's browsing time period and the keyword information of the user's historical browsing content, the first push index of the content keywords is obtained.

[0026] Preferably, the step of obtaining the first push index of content keywords based on user browsing time period information and user historical browsing content keyword information specifically includes:

[0027] Based on the user's browsing time period information and the keyword information of the user's historical browsing content, the keyword information of the user's historical browsing time period is obtained, which includes the frequency information of each keyword within each user's browsing time period;

[0028] Based on the keyword information of the user's historical browsing time period, obtain the first push index of the content keywords;

[0029] The formula for calculating the first push index of content keywords is as follows:

[0030]

[0031] In the formula, The first push index for the xth content keyword. This indicates the content keyword belonging to the first (x) keyword. The number of times a user appears within a browsing time period. Indicates the first The total number of keywords appearing within a user's browsing time period. Indicates the first Let D be the set of user browsing time periods, and n be the total number of user browsing time periods.

[0032] Preferably, obtaining the second push index for content keywords based on user behavior characteristic information and user's historical browsing content keyword information specifically includes:

[0033] Based on the user's historical browsing data, user behavior characteristic information is obtained, including user browsing method information and user historical browsing time information;

[0034] Based on user behavior characteristics, obtain the user's full browsing information and the user's incomplete browsing information;

[0035] Based on the user's full browsing information, obtain the user's full browsing keyword information;

[0036] Based on user behavior characteristics, user's total browsing keyword information, and user's historical browsing content keyword information, obtain the second push index for content keywords;

[0037] The formula for calculating the second push index of content keywords is as follows:

[0038]

[0039] In the formula, The second push index for the xth content keyword. The browsing time of the user's historical browsing content containing the xth keyword. The impact coefficient of content push on user browsing time. The end time of browsing for the user whose historical browsing content contains the xth keyword is specified. The impact coefficient of content push notifications based on the user's browsing end time. Let i be the total time of the i-th user's historical browsing content. This represents the user's total browsing impact coefficient for the x-th content keyword, where if ,but ,like ,but , where m is the total number of historical browsing contents of the user containing the x-th content keyword.

[0040] Preferably, obtaining the keyword user push coefficient based on the user's historical search information specifically includes:

[0041] Based on the user's historical search information, obtain search time information and search keyword information;

[0042] Based on the search time information, obtain the search time segment information according to the time segmentation.

[0043] Based on the user's historical search information, obtain the user's historical search content information;

[0044] Based on search time segmentation information, user historical search content information, and search keyword information, obtain the keyword user push coefficient;

[0045] The formula for calculating the keyword user push coefficient is as follows:

[0046]

[0047] In the formula, Let x be the keyword user push coefficient for the x-th content keyword. This represents the weight coefficient for the j-th search time period. This represents the search ranking index of the x-th content keyword in the j-th search time period. The value represents the user browsing time of the s-th user's historical search content for the x-th content keyword, h is the total number of segments in the search time period, and g is the total number of user's historical search content for the x-th content keyword.

[0048] Preferably, the precise filtering of content to be pushed based on the content push matching index specifically includes:

[0049] Based on the information of the content to be pushed, the first push index of the content keywords, the second push index of the content keywords, and the keyword user push coefficient, obtain the content push matching index;

[0050] Based on the need for precise content filtering, obtain the content push matching index threshold;

[0051] Based on the content push matching index and the content push matching index threshold, determine whether the content push matching index is lower than the content push matching index threshold. If it is, the content does not meet the push standard and is filtered. If not, the content meets the push standard.

[0052] The formula for calculating the content recommendation matching index is as follows:

[0053]

[0054] In the formula, K is the content matching index of the content to be pushed. This indicates the user's browsing time period to which the content was pushed. The first push index for the xth content keyword. The second push index for the xth content keyword. Let be the keyword user push coefficient for the x-th content keyword, and w be the total number of content keywords to be pushed.

[0055] Furthermore, a content precision filtering system based on big data behavioral feature analysis is proposed to implement the filtering method described above, including:

[0056] The main control module is used to obtain user historical browsing time classification information based on the user's historical browsing time, obtain browsing content attribute change information based on the user's basic browsing time period information and browsing content attribute keyword information, obtain user browsing time period information based on browsing content attribute change information and browsing content attribute change threshold, obtain search time segment information based on search time information and time segmentation, obtain content push matching index based on content to be pushed information, first push index of content keywords, second push index of content keywords and keyword user push coefficient, and determine whether the content push matching index is lower than the content push matching index threshold based on the content push matching index and the content push matching index threshold, so as to accurately filter the content to be pushed;

[0057] The information acquisition module is used to acquire user historical browsing data, user historical browsing time information, user historical browsing time information, and user historical browsing content information. Based on the user historical browsing data, it acquires keyword information of the user historical browsing content, acquires attribute keyword information of the browsing content based on the keyword information of the user historical browsing content, acquires user historical search information based on the user historical browsing data, and transmits it to the evaluation module.

[0058] The evaluation module is used to obtain keyword information of the user's historical browsing time period based on the user's browsing time period information and the keyword information of the user's historical browsing content; to obtain the first push index of the content keyword based on the keyword information of the user's historical browsing time period; to obtain the second push index of the content keyword based on the user's behavioral characteristics information, the user's total browsing keyword information and the keyword information of the user's historical browsing content; and to obtain the keyword user push coefficient based on the search time segment information, the user's historical search content information and search keyword information.

[0059] The display module interacts with the main control module and is used to display the first push index of content keywords, the second push index of content keywords, the keyword user push coefficient, and the content push matching index, and output the filtering results of the content to be pushed.

[0060] Optionally, the main control module specifically includes:

[0061] The control unit is used to obtain a content push matching index based on the content to be pushed information, the first push index of the content keyword, the second push index of the content keyword, and the keyword user push coefficient; and to determine whether the content push matching index is lower than the content push matching index threshold based on the content push matching index and the content push matching index threshold, so as to accurately filter the content to be pushed.

[0062] An information receiving unit, which interacts with the information acquisition module and the evaluation module, is used to receive data and transmit it to the data analysis unit.

[0063] The data analysis unit is used to obtain user historical browsing time classification information based on the user's historical browsing time, obtain browsing content attribute change information based on the user's basic browsing time period information and browsing content attribute keyword information, obtain user browsing time period information based on browsing content attribute change information and browsing content attribute change threshold, and obtain search time segment information based on search time information and time segmentation.

[0064] Optionally, the information acquisition module specifically includes:

[0065] An information acquisition unit is used to acquire user historical browsing data, user historical browsing time information, user historical browsing time information, and user historical browsing content information.

[0066] The information processing unit is used to obtain keyword information of the user's historical browsing content based on the user's historical browsing data, obtain attribute keyword information of the browsing content based on the keyword information of the user's historical browsing content, obtain the user's historical search information based on the user's historical browsing data, and transmit it to the evaluation module.

[0067] Optionally, the evaluation module specifically includes:

[0068] The first evaluation unit is used to obtain the keyword information of the user's historical browsing time period based on the user's browsing time period information and the keyword information of the user's historical browsing content, and to obtain the first push index of the content keyword based on the keyword information of the user's historical browsing time period.

[0069] The second evaluation unit is used to obtain the second push index of content keywords based on user behavior feature information, user full browsing keyword information and user historical browsing content keyword information.

[0070] The search evaluation unit is used to obtain the keyword user push coefficient based on search time segmentation information, user historical search content information, and search keyword information.

[0071] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0072] This invention proposes a method and system for precise content filtering based on big data behavioral feature analysis. It analyzes users' browsing preferences at different times through a first content keyword push index, analyzes users' behavioral characteristics when browsing content through a second content keyword push index, accurately evaluates users' historical search keywords through a keyword user push coefficient, and precisely filters the content to be pushed through a content push matching index, thereby improving user experience and optimizing content dissemination. Attached Figure Description

[0073] Figure 1 This is a flowchart of the content precision filtering method based on big data behavioral feature analysis proposed in this invention;

[0074] Figure 2 This is a flowchart of the process for obtaining the first push index of content keywords in this invention;

[0075] Figure 3 This is a flowchart of the process for obtaining the second push index of content keywords in this invention;

[0076] Figure 4 This is a block diagram of the content precision filtering system based on big data behavioral feature analysis proposed in this invention. Detailed Implementation

[0077] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0078] Reference Figure 1-3 As shown in the embodiment of the present invention, the content precision filtering method based on big data behavioral feature analysis includes:

[0079] Obtain user's historical browsing data, which includes user's historical browsing time information, user's historical browsing time information, and user's historical browsing content information;

[0080] Based on the user's historical browsing data, obtain the keyword information of the user's historical browsing content. The keyword information of the user's historical browsing content includes the keyword information of browsing content features and the keyword information of browsing content attributes.

[0081] Based on the user's historical browsing data, obtain the user's historical browsing time classification information;

[0082] Based on the user's historical browsing time category information and the user's historical browsing content keyword information, obtain the first push index of content keywords;

[0083] Specifically, based on user browsing history time category information and user browsing history content keyword information, the first push index of content keywords is obtained, including:

[0084] Based on the user's browsing history data, obtain the user's browsing history time information;

[0085] Based on the user's historical browsing time information, obtain the user's basic browsing time period information;

[0086] Based on the keyword information of the user's historical browsing content, obtain the keyword information of the browsing content attributes;

[0087] Based on the user's basic browsing time period information and the keyword information of the browsing content attributes, the browsing content attribute change information is obtained. The browsing content attribute change information includes the proportion change information of the user's historical browsing content with different attributes within the basic browsing time period for each user.

[0088] Based on the need for precise content filtering, obtain the threshold for changes in the attributes of the browsed content;

[0089] Based on the information on changes in the attributes of the browsed content and the threshold for changes in the attributes of the browsed content, information on the user's browsing time period is obtained;

[0090] Based on the user's browsing time period and the keyword information of the user's historical browsing content, the first push index of the content keywords is obtained.

[0091] Specifically, based on user browsing time period information and user historical browsing keyword information, the first push index of content keywords is obtained, which includes:

[0092] Based on the user's browsing time period information and the keyword information of the user's historical browsing content, the keyword information of the user's historical browsing time period is obtained, which includes the frequency information of each keyword within each user's browsing time period;

[0093] Based on the keyword information of the user's historical browsing time period, obtain the first push index of the content keywords;

[0094] The formula for calculating the first push index of content keywords is as follows:

[0095]

[0096] In the formula, The first push index for the xth content keyword. This indicates the content keyword belonging to the first (x) keyword. The number of times a user appears within a browsing time period. Indicates the first The total number of keywords appearing within a user's browsing time period. Indicates the first Let D be the set of user browsing time periods, and n be the total number of user browsing time periods.

[0097] In this solution, the user's basic browsing time period information is obtained through the user's historical browsing time information; the browsing content attribute keyword information is obtained based on the user's historical browsing content keyword information; the user's browsing time period information is obtained based on the browsing content attribute change information and the browsing content attribute change threshold; and the first push index of the content keyword is obtained based on the user's browsing time period information and the user's historical browsing content keyword information.

[0098] In this embodiment, the content browsed by users within a day is analyzed on a cyclical basis. It is understood that users' preferences for different content attributes vary at different times of the day. For example, users prefer browsing text information at midday and video information at night. Keywords for browsing content attributes include text, video, TV series, and live streams. By using information on changes in browsing content attributes and thresholds for these changes, the day is divided into different time periods. For instance, during the daytime, users browse the largest proportion of text content, while at night, they browse the largest proportion of video content. Therefore, the day can be divided into daytime and nighttime periods. If the change in the proportion of browsing content attributes does not exceed the threshold, that content attribute is not assigned a time period, thus preventing users from browsing too little content from affecting content filtering.

[0099] It is understandable that even for the same keyword, users' browsing preferences for that keyword will differ at different times.

[0100] Based on users' historical browsing data, obtain user behavior characteristic information;

[0101] Based on user behavior characteristics and user browsing history keyword information, obtain the second push index for content keywords;

[0102] Specifically, based on user behavior characteristics and historical browsing keyword information, a second push index for content keywords is obtained, including:

[0103] Based on the user's historical browsing data, user behavior characteristic information is obtained, including user browsing method information and user historical browsing time information;

[0104] Based on user behavior characteristics, obtain the user's full browsing information and the user's incomplete browsing information;

[0105] Based on the user's full browsing information, obtain the user's full browsing keyword information;

[0106] Based on user behavior characteristics, user's total browsing keyword information, and user's historical browsing content keyword information, obtain the second push index for content keywords;

[0107] The formula for calculating the second push index of content keywords is as follows:

[0108]

[0109] In the formula, The second push index for the xth content keyword. The browsing time of the user's historical browsing content containing the xth keyword. The impact coefficient of content push on user browsing time. The end time of browsing for the user whose historical browsing content contains the xth keyword is specified. The impact coefficient of content push notifications based on the user's browsing end time. Let i be the total time of the i-th user's historical browsing content. This represents the user's total browsing impact coefficient for the x-th content keyword, where if ,but ,like ,but , where m is the total number of historical browsing contents of the user containing the x-th content keyword.

[0110] In this solution, user behavior feature information is obtained through user's historical browsing data. Based on the user behavior feature information, the user's full browsing information and incomplete browsing information are obtained. Based on the user's full browsing information, the user's full browsing keyword information is obtained. Based on the user behavior feature information, the user's full browsing keyword information and the user's historical browsing content keyword information, the second push index of content keywords is obtained, and the user's behavioral characteristics when browsing content are analyzed.

[0111] Understandably, the way users browse content and the time they spend browsing reflect their preference for content with different keywords. For content that is fully viewed, content containing the same keywords is more likely to be viewed by users. However, for content that is not fully viewed, even if users spend the same amount of time browsing, different browsing methods still indicate different user preferences. For example, for video content A and video content B, users may spend the same amount of time browsing, but for video content A, a user may close the video halfway through, while for video content B, a user may skip to the end of video content B and spend half of the video content B's length. In this case, the user's preference for video content A and video content B is completely different.

[0112] Based on the user's historical browsing data, the user's historical search information is obtained, which includes search keyword information and search time information;

[0113] Based on users' historical search information, obtain the keyword user push coefficient;

[0114] Specifically, based on users' historical search information, the keyword user recommendation coefficient is obtained, including:

[0115] Based on the user's historical search information, obtain search time information and search keyword information;

[0116] Based on the search time information, obtain the search time segment information according to the time segmentation.

[0117] Based on the user's historical search information, obtain the user's historical search content information;

[0118] Based on search time segmentation information, user historical search content information, and search keyword information, obtain the keyword user push coefficient;

[0119] The formula for calculating the keyword user push coefficient is as follows:

[0120]

[0121] In the formula, Let x be the keyword user push coefficient for the x-th content keyword. This represents the weight coefficient for the j-th search time period. This represents the search ranking index of the x-th content keyword in the j-th search time period. The value represents the user browsing time of the s-th user's historical search content for the x-th content keyword, h is the total number of segments in the search time period, and g is the total number of user's historical search content for the x-th content keyword.

[0122] In this solution, search time and keyword information are obtained through users' historical search information. Based on the search time information, search time segmentation information is obtained, such as dividing the search time into recent search time period, recent search time period, and historical search time period. Based on the search time segmentation information, users' historical search content information, and search keyword information, keyword user push coefficient is obtained. It can be understood that different search times for keywords reflect the degree of user preference for those keywords. Through the keyword user push coefficient, the user's historical search keywords can be accurately evaluated.

[0123] Obtain information about the content to be pushed, including keyword information for the content to be pushed;

[0124] Based on the information of the content to be pushed, the first push index of the content keywords, the second push index of the content keywords, and the keyword user push coefficient, obtain the content push matching index;

[0125] Based on the content matching index, the content to be pushed is accurately filtered.

[0126] Specifically, based on the content matching index, the content to be pushed is precisely filtered, including:

[0127] Based on the information of the content to be pushed, the first push index of the content keywords, the second push index of the content keywords, and the keyword user push coefficient, obtain the content push matching index;

[0128] Based on the need for precise content filtering, obtain the content push matching index threshold;

[0129] Based on the content push matching index and the content push matching index threshold, determine whether the content push matching index is lower than the content push matching index threshold. If it is, the content does not meet the push standard and is filtered. If not, the content meets the push standard.

[0130] The formula for calculating the content recommendation matching index is as follows:

[0131]

[0132] In the formula, K is the content matching index of the content to be pushed. This indicates the user's browsing time period to which the content was pushed. The first push index for the xth content keyword. The second push index for the xth content keyword. Let be the keyword user push coefficient for the x-th content keyword, and w be the total number of content keywords to be pushed.

[0133] In this solution, a content push matching index is obtained by using the content information to be pushed, the first push index of content keywords, the second push index of content keywords, and the keyword user push coefficient. Based on the content push matching index and the content push matching index threshold, it is determined whether the content push matching index is lower than the content push matching index threshold. By analyzing the keywords of the content to be pushed, the user's preference for each keyword is accurately assessed, enabling precise filtering of the content to be pushed. This improves user experience and optimizes the content dissemination effect. It can be understood that content keywords include content feature keywords and content attribute keywords. Content feature keywords include the specific subject matter of the content, such as medical, food, lifestyle, entertainment, etc., while content attribute keywords include the specific attributes of the content, such as text, video, live broadcast, TV series, movie, etc.

[0134] Reference Figure 4 As shown, further, combining the above-mentioned content precision filtering method based on big data behavioral feature analysis, a content precision filtering system based on big data behavioral feature analysis is proposed, including:

[0135] The main control module is used to obtain user historical browsing time classification information based on the user's historical browsing time, obtain browsing content attribute change information based on the user's basic browsing time period information and browsing content attribute keyword information, obtain user browsing time period information based on browsing content attribute change information and browsing content attribute change threshold, obtain search time segment information based on search time information and time segmentation, obtain content push matching index based on content to be pushed information, first push index of content keywords, second push index of content keywords and keyword user push coefficient, and determine whether the content push matching index is lower than the content push matching index threshold based on the content push matching index and the content push matching index threshold, so as to accurately filter the content to be pushed;

[0136] The information acquisition module is used to acquire user historical browsing data, user historical browsing time information, user historical browsing time information, and user historical browsing content information. Based on the user historical browsing data, it acquires keyword information of the user historical browsing content, acquires attribute keyword information of the browsing content based on the keyword information of the user historical browsing content, acquires user historical search information based on the user historical browsing data, and transmits it to the evaluation module.

[0137] The evaluation module is used to obtain keyword information of the user's historical browsing time period based on the user's browsing time period information and the keyword information of the user's historical browsing content; to obtain the first push index of the content keyword based on the keyword information of the user's historical browsing time period; to obtain the second push index of the content keyword based on the user's behavioral characteristics information, the user's total browsing keyword information and the keyword information of the user's historical browsing content; and to obtain the keyword user push coefficient based on the search time segment information, the user's historical search content information and search keyword information.

[0138] The display module interacts with the main control module and is used to display the first push index of content keywords, the second push index of content keywords, the keyword user push coefficient, and the content push matching index, and output the filtering results of the content to be pushed.

[0139] The main control module specifically includes:

[0140] The control unit is used to obtain a content push matching index based on the content to be pushed information, the first push index of the content keyword, the second push index of the content keyword, and the keyword user push coefficient; and to determine whether the content push matching index is lower than the content push matching index threshold based on the content push matching index and the content push matching index threshold, so as to accurately filter the content to be pushed.

[0141] An information receiving unit, which interacts with the information acquisition module and the evaluation module, is used to receive data and transmit it to the data analysis unit.

[0142] The data analysis unit is used to obtain user historical browsing time classification information based on the user's historical browsing time, obtain browsing content attribute change information based on the user's basic browsing time period information and browsing content attribute keyword information, obtain user browsing time period information based on browsing content attribute change information and browsing content attribute change threshold, and obtain search time segment information based on search time information and time segmentation.

[0143] The information acquisition module specifically includes:

[0144] An information acquisition unit is used to acquire user historical browsing data, user historical browsing time information, user historical browsing time information, and user historical browsing content information.

[0145] The information processing unit is used to obtain keyword information of the user's historical browsing content based on the user's historical browsing data, obtain attribute keyword information of the browsing content based on the keyword information of the user's historical browsing content, obtain the user's historical search information based on the user's historical browsing data, and transmit it to the evaluation module.

[0146] The evaluation module specifically includes:

[0147] The first evaluation unit is used to obtain the keyword information of the user's historical browsing time period based on the user's browsing time period information and the keyword information of the user's historical browsing content, and to obtain the first push index of the content keyword based on the keyword information of the user's historical browsing time period.

[0148] The second evaluation unit is used to obtain the second push index of content keywords based on user behavior feature information, user full browsing keyword information and user historical browsing content keyword information.

[0149] The search evaluation unit is used to obtain the keyword user push coefficient based on search time segmentation information, user historical search content information, and search keyword information.

[0150] In summary, the advantages of this invention are as follows: By acquiring keyword information of previously viewed content through user historical browsing data, and by acquiring a first content keyword push index based on user historical browsing time classification information and keyword information, the user's browsing content preferences at different time periods are analyzed. Furthermore, by acquiring a second content keyword push index based on user behavior characteristics and keyword information, the user's behavioral characteristics while browsing content are analyzed. Finally, by acquiring a keyword user push coefficient based on user historical search information, the user's historical search keywords are accurately evaluated. Finally, by acquiring a content push matching index based on the content to be pushed, the first content keyword push index, the second content keyword push index, and the keyword user push coefficient, the content to be pushed is precisely filtered, thereby improving user experience and optimizing content dissemination effects.

[0151] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for accurate content screening based on big data behavior feature analysis, characterized in that: include: Acquire user historical browsing data, wherein the user historical browsing data includes user historical browsing moment information, user historical browsing time information, and user historical browsing content information; According to the user's historical browsing data, the keyword information of the user's historical browsing content is obtained, wherein the keyword information of the user's historical browsing content includes the keyword information of the browsing content feature and the keyword information of the browsing content attribute; According to the user's historical browsing data, based on the user's historical browsing time, obtain the user's historical browsing time classification information; According to the user's historical browsing time classification information and the user's historical browsing content keyword information, obtain the content keyword first push index; Obtain user behavior feature information based on user historical browsing data; According to the user behavior feature information and the user's historical browsing content keyword information, obtain the content keyword second push index; Acquire user historical search information according to user historical browsing data, wherein the user historical search information includes search keyword information and search time information; Obtain keyword user push coefficient based on user historical search information; Acquire information about content to be pushed, wherein the information about content to be pushed includes keyword information about the content to be pushed; Obtaining a content push matching index based on the content information to be pushed, the first push index of the content keyword, the second push index of the content keyword, and the keyword user push coefficient; According to the content push matching index, the pushed content can be accurately screened.

2. The method for accurate content screening based on big data behavior feature analysis according to claim 1 is characterized in that: The step of obtaining the first content keyword push index according to the user's historical browsing time classification information and the user's historical browsing content keyword information specifically includes: According to the user's historical browsing data, obtain the user's historical browsing time information; According to the user's historical browsing time information, obtain the user's browsing basic time period information; Obtaining keyword information of browsing content attributes based on keyword information of user's historical browsing content; According to the basic browsing time period information of the user and the keyword information of the browsing content attribute, the browsing content attribute change information is obtained, wherein the browsing content attribute change information includes the ratio change information of the user's historical browsing content with different attributes within the basic browsing time period of each user; Based on the needs of accurate content screening, obtain the threshold value of browsing content attribute changes; Obtaining user browsing time period information according to browsing content attribute change information and browsing content attribute change threshold; According to the user's browsing time period information and the user's historical browsing content keyword information, the content keyword first push index is obtained.

3. The method for accurate content screening based on big data behavior feature analysis according to claim 2 is characterized in that: The step of obtaining the first content keyword push index according to the user browsing time period information and the user's historical browsing content keyword information specifically includes: According to the user browsing time period information and the user's historical browsing content keyword information, the user's historical browsing time period keyword information is obtained, wherein the user's historical browsing time period keyword information includes the frequency information of each keyword in each user's browsing time period; Obtain the first push index of content keywords based on the keyword information of the user's historical browsing time period; Among them, the calculation formula for the first push index of content keywords is: In the formula, The first push index of the x-th content keyword. Indicates the xth content keyword belongs to The number of times a user appears during a browsing period. Indicates The total number of keywords that appear during a user's browsing time period. Indicates user browsing time periods, D is the set of user browsing time periods, and n is the total number of user browsing time periods.

4. The method for accurate content screening based on big data behavior feature analysis according to claim 1 is characterized in that: The obtaining of the second content keyword push index according to the user behavior feature information and the user's historical browsing content keyword information specifically includes: Obtaining user behavior feature information based on user historical browsing data, wherein the user behavior feature information includes user browsing mode information and user historical browsing time information; According to the user behavior feature information, obtain the user's full browsing information and the user's incomplete browsing information; According to the user's full browsing information, obtain the user's full browsing keyword information; Obtain the second push index of content keywords based on user behavior feature information, user full browsing keyword information, and user historical browsing content keyword information; Among them, the calculation formula for the second push index of content keywords is: In the formula, The second push index of the xth content keyword. is the user browsing time of the i-th user's historical browsing content containing the x-th content keyword, The content push impact coefficient for user browsing time, The end time of the user browsing the i-th user's historical browsing content containing the x-th content keyword, The impact coefficient of content push for the end time of user browsing. is the total time of historical browsing content by the i-th user, represents the user's full browsing influence coefficient of the xth content keyword, where ,but ,like ,but , m is the total number of historically browsed contents by users that contain the xth content keyword.

5. The method for accurate content screening based on big data behavior feature analysis according to claim 1 is characterized in that: The method of obtaining a keyword user push coefficient according to the user's historical search information specifically includes: According to the user's historical search information, obtain the search time information and search keyword information; According to the search time information, based on the time segment, the search time segment information is obtained; According to the user's historical search information, obtain the user's historical search content information; Obtain keyword user push coefficient based on search time segment information, user historical search content information and search keyword information; Among them, the calculation formula for the keyword user push coefficient is: In the formula, is the keyword user push coefficient of the xth content keyword, represents the weight coefficient of the jth search time period, It represents the search order index of the xth content keyword in the jth search time period. represents the user browsing time of the sth user's historical search content for the xth content keyword, h is the total number of search time periods, and g is the total number of user's historical search content for the xth content keyword.

6. The method for accurate content screening based on big data behavior feature analysis according to claim 1 is characterized in that: The accurate screening of the content to be pushed according to the content push matching index specifically includes: Obtaining a content push matching index based on the content information to be pushed, the first push index of the content keyword, the second push index of the content keyword, and the keyword user push coefficient; Based on the needs of accurate content screening, obtain the content push matching index threshold; According to the content push matching index and the content push matching index threshold, determine whether the content push matching index is lower than the content push matching index threshold. If so, the content does not meet the push standard and the content is screened. If not, the content meets the push standard. The calculation formula of content push matching index is: Where K is the content push matching index of the content to be pushed, Indicates the user browsing time period to which the content is pushed. The first push index of the x-th content keyword. The second push index of the xth content keyword. is the keyword user push coefficient of the xth content keyword, and w is the total number of content keywords to be pushed.

7. A content accurate screening system based on big data behavior feature analysis, used to implement the screening method according to any one of claims 1 to 6, characterized in that: include: A main control module, the main control module is used to obtain user historical browsing time classification information based on user historical browsing data and user historical browsing time, obtain browsing content attribute change information based on user browsing basic time period information and browsing content attribute keyword information, obtain user browsing time period information based on browsing content attribute change information and browsing content attribute change threshold, obtain search time segment information based on time segmentation according to search time information, obtain content push matching index according to content information to be pushed, content keyword first push index, content keyword second push index and keyword user push coefficient, determine whether the content push matching index is lower than the content push matching index threshold according to the content push matching index and the content push matching index threshold, and accurately screen the content to be pushed; An information acquisition module, the information acquisition module is used to acquire user historical browsing data, user historical browsing moment information, user historical browsing time information and user historical browsing content information, acquire user historical browsing content keyword information based on user historical browsing data, acquire browsing content attribute keyword information based on user historical browsing content keyword information, acquire user historical search information based on user historical browsing data, and transmit it to the evaluation module; An evaluation module, the evaluation module is used to obtain the user's historical browsing time period keyword information according to the user's browsing time period information and the user's historical browsing content keyword information, obtain the content keyword first push index according to the user's historical browsing time period keyword information, obtain the content keyword second push index according to the user's behavior feature information, the user's full browsing keyword information and the user's historical browsing content keyword information, and obtain the keyword user push coefficient according to the search time segment information, the user's historical search content information and the search keyword information; The display module interacts with the main control module and is used to display the first push index of the content keyword, the second push index of the content keyword, the keyword user push coefficient and the content push matching index, and output and display the screening results of the content to be pushed.

8. The content accurate screening system based on big data behavior feature analysis according to claim 7 is characterized in that: The main control module specifically includes: A control unit, the control unit is used to obtain a content push matching index according to the content information to be pushed, a first content keyword push index, a second content keyword push index, and a keyword user push coefficient, and to determine whether the content push matching index is lower than the content push matching index threshold according to the content push matching index and the content push matching index threshold, so as to accurately screen the content to be pushed; An information receiving unit, which interacts with the information acquisition module and the evaluation module to receive data and transmit it to the data analysis unit; A data analysis unit, wherein the data analysis unit is used to obtain classification information of user's historical browsing moments based on user's historical browsing data, obtain browsing content attribute change information based on user's browsing basic time period information and browsing content attribute keyword information, obtain user's browsing time period information based on browsing content attribute change information and browsing content attribute change threshold, and obtain search time segment information based on search time information and time segmentation.

9. The content accurate screening system based on big data behavior feature analysis according to claim 7 is characterized in that: The information acquisition module specifically includes: An information acquisition unit, the information acquisition unit is used to acquire user historical browsing data, user historical browsing moment information, user historical browsing time information and user historical browsing content information; An information processing unit is used to obtain keyword information of the user's historical browsing content based on the user's historical browsing data, obtain keyword information of browsing content attributes based on the user's historical browsing content keyword information, obtain user's historical search information based on the user's historical browsing data, and transmit it to the evaluation module.

10. The content accurate screening system based on big data behavior feature analysis according to claim 7 is characterized in that: The evaluation module specifically includes: A first evaluation unit, the first evaluation unit is used to obtain the keyword information of the user's historical browsing time period according to the user's browsing time period information and the keyword information of the user's historical browsing content, and obtain the first content keyword push index according to the keyword information of the user's historical browsing time period; A second evaluation unit, the second evaluation unit is used to obtain a content keyword second push index according to user behavior feature information, user full browsing keyword information and user historical browsing content keyword information; The search evaluation unit is used to obtain a keyword user push coefficient according to search time segment information, user history search content information and search keyword information.