A method for intelligently pushing information data based on customized windows

Abstract

The present invention relates to the field of information processing technology, and in particular to a method for intelligently pushing information data based on a customized window. The method comprises the following steps: obtaining information platform behavior log record data and performing behavior mining analysis to obtain information platform user browsing behavior data and information platform user search behavior data; constructing a user behavior preference matrix for the information platform user browsing behavior data and the information platform user search behavior data to obtain an information platform user behavior preference feature matrix; performing browsing information screening processing and context association perception analysis on the information platform behavior log record data to obtain an information context association perception sequence data set; performing information push rule mining analysis and intelligent matching push processing based on the information platform user behavior preference feature matrix and the information context association perception sequence data set to obtain an information intelligent matching push sequence set. The present invention can realize the intelligent push of information content that meets user preference requirements.

Description

Technical Field

[0001] The present invention relates to the field of information processing technology, and in particular to an information data intelligent push method based on a customized window. Background Art

[0002] In the current era of information explosion, people are faced with the trouble of massive information. How to filter out valuable content from a large amount of information has become a challenge. With the rapid development of information technology, people's demand for personalized and accurate information push is increasing, and data intelligent push has become the main way for people to obtain personalized information. At present, information push mainly relies on users' historical behavior data and recommendation algorithms. It takes advantage of modern technologies such as artificial intelligence, big data analysis and machine learning. By real-time monitoring and analysis of user behavior, it continuously optimizes the push algorithm, making the pushed information more in line with users' actual needs and interests. At the same time, the adoption of customized windows also allows users to freely set the push content and method according to their preferences and needs. However, traditional information recommendation algorithms can often only make recommendations based on users' click behavior or simple interest tags, lacking a deep understanding of user interests and accurate prediction. Summary of the invention

[0003] Based on this, it is necessary for the present invention to provide an intelligent information data push method based on a customized window to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a method for intelligently pushing information data based on a customized window includes the following steps:

[0005] Step S1: obtaining information platform behavior log record data, and performing user browsing behavior and search behavior mining and analysis on the information platform behavior log record data to obtain information platform user browsing behavior data and information platform user search behavior data; constructing a user behavior preference matrix for the information platform user browsing behavior data and the information platform user search behavior data to obtain an information platform user behavior preference feature matrix;

[0006] Step S2: Perform browsing information screening on the information platform behavior log record data to obtain a user browsing related information data set; perform context association perception analysis on the user browsing related information data set to obtain an information context association perception sequence data set;

[0007] Step S3: performing information push rule mining and analysis based on the information platform user behavior preference feature matrix and the information context association perception sequence data set to obtain the information platform information push rule data set; performing customized window construction based on the information platform information push rule data set to generate the information platform information push customized window;

[0008] Step S4: Acquire the information data set to be pushed, and perform intelligent matching and pushing processing on the information data set to be pushed according to the information platform information push customization window to obtain an information intelligent matching push sequence set.

[0009] The present invention first obtains the behavior log record data of the information platform, and can obtain various behavior data of users on the information platform, such as browsing, searching, clicking, etc. These data record the activity trajectory and behavior pattern of users on the platform. Obtaining these data is crucial for understanding user behavior, analyzing user preferences, and optimizing platform services. By mining and analyzing the user browsing behavior of the information platform behavior log record data, it is possible to deeply understand the browsing behavior of users on the information platform, including the pages visited, the dwell time, the clicked links, and other information. Such an analysis and processing process can help discover the user's interests and preferred areas, and then provide users with more personalized content recommendations and services that meet their needs. In addition, by mining and analyzing the user search behavior of the information platform behavior log record data, it is possible to understand the user's search behavior on the information platform, including search keywords, search frequency, search result clicks, etc. Such an analysis and processing process helps to understand the user's needs and intentions, thereby providing users with more accurate and effective search suggestions and results display. At the same time, by constructing a user behavior preference matrix for the information platform user browsing behavior data and the information platform user search behavior data, the information platform user browsing behavior data and the user search behavior data can be combined to construct a user behavior preference matrix. Through such matrix construction, the user's behavior preferences and interest characteristics on the information platform can be fully reflected, thereby improving the in-depth understanding of user interests and providing an important data basis and support for the subsequent customized window process. Secondly, by filtering the browsing information of the information platform behavior log record data, the user's browsing related information data set can be extracted. This step can remove irrelevant information by cleaning, filtering and classifying the behavior log data to extract the user's browsing behavior related data set. This data set contains the user's browsing records on the information platform, thus providing a data basis for subsequent context association mining analysis. By performing context association perception analysis on the user's browsing related information data set, an information context association perception sequence data set can be obtained, in which each information is arranged according to the connection relationship of its corresponding context information. This sequence data set reflects that the information data in the user's browsing related information data set is sorted according to the context information connection relationship, thus providing an important basis for the subsequent customized window construction and content push. Then, by combining the information platform user behavior preference feature matrix and the information context association perception sequence data set to conduct information push rule mining and analysis to analyze user behavior preferences and context information, we can discover user preference patterns and behavior regularities, and then determine the initial information push rules. These rules can help push information of interest to users more accurately, thereby improving the accuracy and effectiveness of information push.A customized window is also constructed according to the information platform information push rule data set to generate a personalized information push interface. The information push rules generated according to user behavior preferences and context information can dynamically adjust the content and form of information push to present the most relevant and attractive information to users. This step of customizing the window construction can provide users with a better information browsing experience, thereby enhancing users' stickiness and loyalty to the information platform. Finally, by obtaining the information data set to be pushed, including but not limited to various types of information content such as articles, news, and advertisements, this step can provide users with diversified and personalized information content by collecting the information data set to be pushed to meet their different needs and interests. By intelligently matching and pushing the information data set to be pushed according to the information platform information push customized window, it is ensured that the information pushed to the user is closer to his interests and needs, and the best push time is determined according to factors such as the user's active time and preferences. This step can ensure that users receive information push at the most appropriate time, increase users' attention to information and click-through rate, and the intelligent timed push processing process can also reduce users' interference with unnecessary information, thereby improving the accurate prediction and effect of information push. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0011] Figure 1 It is a schematic diagram of the steps of the method for intelligently pushing information data based on a customized window of the present invention;

[0012] Figure 2 for Figure 1 Detailed step flow diagram of step S1;

[0013] Figure 3 for Figure 2 Detailed step flow chart of step S12 in FIG. DETAILED DESCRIPTION

[0014] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0015] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0016] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0017] To achieve this, please refer to Figures 1 to 3 The present invention provides a method for intelligently pushing information data based on a customized window, the method comprising the following steps:

[0018] Step S1: obtaining information platform behavior log record data, and performing user browsing behavior and search behavior mining and analysis on the information platform behavior log record data to obtain information platform user browsing behavior data and information platform user search behavior data; constructing a user behavior preference matrix for the information platform user browsing behavior data and the information platform user search behavior data to obtain an information platform user behavior preference feature matrix;

[0019] Step S2: Perform browsing information screening on the information platform behavior log record data to obtain a user browsing related information data set; perform context association perception analysis on the user browsing related information data set to obtain an information context association perception sequence data set;

[0020] Step S3: performing information push rule mining and analysis based on the information platform user behavior preference feature matrix and the information context association perception sequence data set to obtain the information platform information push rule data set; performing customized window construction based on the information platform information push rule data set to generate the information platform information push customized window;

[0021] Step S4: Acquire the information data set to be pushed, and perform intelligent matching and pushing processing on the information data set to be pushed according to the information platform information push customization window to obtain an information intelligent matching push sequence set.

[0022] In the embodiment of the present invention, please refer to Figure 1 FIG. 1 is a flow chart of the steps of the method for intelligently pushing information data based on a customized window of the present invention. In this example, the method for intelligently pushing information data based on a customized window includes the following steps:

[0023] Step S1: obtaining information platform behavior log record data, and performing user browsing behavior and search behavior mining and analysis on the information platform behavior log record data to obtain information platform user browsing behavior data and information platform user search behavior data; constructing a user behavior preference matrix for the information platform user browsing behavior data and the information platform user search behavior data to obtain an information platform user behavior preference feature matrix;

[0024] The embodiment of the present invention first determines the data source of the behavior log record of the information platform, including server logs, database records, etc., and extracts the data related to user behavior from the log record by designing and executing a data extraction program to obtain various behavior data of users on the information platform, such as browsing, searching, clicking, etc., thereby obtaining the behavior log record data of the information platform. At the same time, by using the corresponding behavior analysis method, the user browsing behavior and user search behavior in the browsing record data of the information platform user are mined and analyzed to find out the key nodes and conversion paths in the user's browsing conversion process and understand the user's search behavior on the information platform, including search keywords, search frequency, search result click status, etc., thereby obtaining the browsing behavior data of the information platform user and the search behavior data of the information platform user. Secondly, by using the feature analysis method, the information platform user browsing behavior data and the information platform user search behavior data are respectively extracted and analyzed to extract the characteristic information of user behavior, such as browsing behavior characteristics such as browsing access frequency, browsing time, browsing habits, browsing time period, and search behavior characteristics such as search keywords, search frequency, and search preference. The browsing behavior characteristics and search behavior characteristics obtained by analysis are analyzed by using the behavior preference index evaluation and analysis method to establish the user behavior preference index by comprehensively analyzing the user's behavior characteristic data, and the behavior preference index of each user is converted into a feature vector. At the same time, the user behavior preference index feature vectors obtained by analysis are classified by using a clustering algorithm or a classification algorithm to classify the behavior preference index feature vectors of similar behavior patterns into the same category, and the classification results are annotated with classification labels to indicate the behavior preference category to which the feature vector belongs. Then, the matrix is ​​integrated and constructed by integrating the feature vectors of each behavior preference index and the corresponding behavior preference index classification label, so that the user's behavior preference feature vectors are integrated and combined according to the classification label, where each row represents a user, and each column represents the user's behavior preference feature vector and its corresponding classification label, and finally the information platform user behavior preference feature matrix is ​​obtained.

[0025] Step S2: Perform browsing information screening on the information platform behavior log record data to obtain a user browsing related information data set; perform context association perception analysis on the user browsing related information data set to obtain an information context association perception sequence data set;

[0026] The embodiment of the present invention collects the behavior log record data of the information platform, including the user's browsing, clicking, searching and other behaviors, and screens the browsing information of the information platform behavior log record data to clean, filter and classify the information platform behavior log record data, and remove irrelevant information to extract the information data set related to the user's browsing behavior, thereby obtaining the user browsing related information data set. At the same time, by using text mining technology (such as natural language processing tools, keyword extraction algorithms, etc.), the information data browsed by each user in the screened user browsing related information data set is processed to identify and extract the most representative and key text information in the user browsing information, which includes titles, abstracts, keywords, etc., and the key text information obtained by mining and analysis is evaluated and analyzed in terms of contextual association relationships to discover the contextual information association rules and patterns between the user browsing information data. Secondly, by combining the contextual information association rules between the various user browsing information data obtained by analysis, the change trend of the contextual information association of the corresponding user browsing information data in the user browsing related information data set is inferred, and the association change trend between the user browsing information data is revealed, including the increase or decrease in the degree of association between the user browsing information data, the change rate, the contextual information associated with the user browsing information data, etc. Then, by combining the contextual information association change trend obtained by analysis, the corresponding user browsing information data in the user browsing related information data set is sorted according to the contextual information connection relationship using the corresponding sorting algorithm (such as sorting based on frequent itemsets, sorting based on confidence, etc.), and finally the information context association perception sequence data set is obtained.

[0027] Step S3: performing information push rule mining and analysis based on the information platform user behavior preference feature matrix and the information context association perception sequence data set to obtain the information platform information push rule data set; performing customized window construction based on the information platform information push rule data set to generate the information platform information push customized window;

[0028] The embodiment of the present invention mines and analyzes the push rules based on the behavior preference characteristic parameters (including click-through rate, browsing history, subscription preference, etc.) in the information platform user behavior preference characteristic matrix and the information data connected and configured in the information context association perception sequence data set to analyze the association between user behavior preference and context information, and discover the user's preference pattern and behavior law, and then determine the initial information push rules, including the time, frequency and content of information push, and optimize the information push rules, thereby generating an information platform information push rule data set. Then, by combining the generated information platform information push rule data set, the parameters and functions of the customized window are designed, including push content, push frequency, push time, etc., to ensure that it can intelligently push information according to the user's behavior preference and the connection relationship between the context information content, and present the most relevant and attractive information to the user, and finally generate an information platform information push customized window.

[0029] Step S4: Acquire the information data set to be pushed, and perform intelligent matching and pushing processing on the information data set to be pushed according to the information platform information push customization window to obtain an information intelligent matching push sequence set.

[0030] The embodiment of the present invention obtains the information data set to be recommended through the information platform or other data sources, including but not limited to various information contents such as articles, news, and advertisements, so as to finally obtain the information data set to be pushed. Secondly, by using the previously designed information platform information push customized window, a matching algorithm is used to match and calculate the information to be pushed in the information data set to be pushed with the historical push information in the customized window, so as to evaluate the matching degree between the information to be pushed and the user interest historical push information in the customized window. At the same time, the corresponding information to be pushed in the information data set to be pushed is sorted according to the matching degree generated by quantification, so as to put the information with higher matching degree in the front, and determine the priority and order of the matching information to be pushed. Then, by using the information platform information push customized window, the information to be pushed sorted in the matching information sequence to be pushed is pushed regularly according to the corresponding push frequency, push time and other push rules, so as to ensure that the user can receive the relevant interest information at the right time, and ensure that the user receives the information push at the most appropriate time, and finally obtain the information intelligent matching push sequence set.

[0031] The present invention first obtains the behavior log record data of the information platform, and can obtain various behavior data of users on the information platform, such as browsing, searching, clicking, etc. These data record the activity trajectory and behavior pattern of users on the platform. Obtaining these data is crucial for understanding user behavior, analyzing user preferences, and optimizing platform services. By mining and analyzing the user browsing behavior of the information platform behavior log record data, it is possible to deeply understand the browsing behavior of users on the information platform, including the pages visited, the dwell time, the clicked links, and other information. Such an analysis and processing process can help discover the user's interests and preferred areas, and then provide users with more personalized content recommendations and services that meet their needs. In addition, by mining and analyzing the user search behavior of the information platform behavior log record data, it is possible to understand the user's search behavior on the information platform, including search keywords, search frequency, search result clicks, etc. Such an analysis and processing process helps to understand the user's needs and intentions, thereby providing users with more accurate and effective search suggestions and results display. At the same time, by constructing a user behavior preference matrix for the information platform user browsing behavior data and the information platform user search behavior data, the information platform user browsing behavior data and the user search behavior data can be combined to construct a user behavior preference matrix. Through such matrix construction, the user's behavior preferences and interest characteristics on the information platform can be fully reflected, thereby improving the in-depth understanding of user interests and providing an important data basis and support for the subsequent customized window process. Secondly, by filtering the browsing information of the information platform behavior log record data, the user's browsing related information data set can be extracted. This step can remove irrelevant information by cleaning, filtering and classifying the behavior log data to extract the user's browsing behavior related data set. This data set contains the user's browsing records on the information platform, thus providing a data basis for subsequent context association mining analysis. By performing context association perception analysis on the user's browsing related information data set, an information context association perception sequence data set can be obtained, in which each information is arranged according to the connection relationship of its corresponding context information. This sequence data set reflects that the information data in the user's browsing related information data set is sorted according to the context information connection relationship, thus providing an important basis for the subsequent customized window construction and content push. Then, by combining the information platform user behavior preference feature matrix and the information context association perception sequence data set to conduct information push rule mining and analysis to analyze user behavior preferences and context information, we can discover user preference patterns and behavior regularities, and then determine the initial information push rules. These rules can help push information of interest to users more accurately, thereby improving the accuracy and effectiveness of information push.A customized window is also constructed according to the information platform information push rule data set to generate a personalized information push interface. The information push rules generated according to user behavior preferences and context information can dynamically adjust the content and form of information push to present the most relevant and attractive information to users. This step of customizing the window construction can provide users with a better information browsing experience, thereby enhancing users' stickiness and loyalty to the information platform. Finally, by obtaining the information data set to be pushed, including but not limited to various types of information content such as articles, news, and advertisements, this step can provide users with diversified and personalized information content by collecting the information data set to be pushed to meet their different needs and interests. By intelligently matching and pushing the information data set to be pushed according to the information platform information push customized window, it is ensured that the information pushed to the user is closer to his interests and needs, and the best push time is determined according to factors such as the user's active time and preferences. This step can ensure that users receive information push at the most appropriate time, increase users' attention to information and click-through rate, and the intelligent timed push processing process can also reduce users' interference with unnecessary information, thereby improving the accurate prediction and effect of information push.

[0032] Preferably, step S1 comprises the following steps:

[0033] Step S11: Obtaining information platform behavior log record data;

[0034] Step S12: performing user browsing behavior mining and analysis on the information platform behavior log record data to obtain information platform user browsing behavior data;

[0035] Step S13: Perform user search behavior mining and analysis on the information platform behavior log record data to obtain information platform user search behavior data;

[0036] Step S14: construct a user behavior preference matrix based on the information platform user browsing behavior data and the information platform user search behavior data to obtain an information platform user behavior preference feature matrix.

[0037] As an embodiment of the present invention, refer to Figure 2 As shown, Figure 1 Detailed step flow diagram of step S1 in FIG. 1 , in this embodiment, step S1 includes the following steps:

[0038] Step S11: Obtaining information platform behavior log record data;

[0039] The embodiment of the present invention first determines the source of the behavior log record data of the information platform, including server logs, database records, etc., and extracts user behavior-related data from the log records by designing and executing a data extraction program to obtain various behavior data of users on the information platform, such as browsing, searching, clicking, etc., and finally obtains the information platform behavior log record data.

[0040] Step S12: performing user browsing behavior mining and analysis on the information platform behavior log record data to obtain information platform user browsing behavior data;

[0041] The embodiment of the present invention first collects the behavior log record data of the information platform, including various operation behaviors of users on the platform, such as browsing, clicking, searching, etc., and screens out the behavior information records related to user browsing according to pre-set screening conditions, including visited pages, browsed content, etc., and analyzes the screened information platform user browsing record data to identify the browsing behavior sequence of users on the information platform, and also analyzes the user's browsing behavior trajectory by using a behavior trajectory analysis method to analyze the user's browsing behavior trajectory and understand the user's browsing habits, preferences, and information acquisition paths and methods on the information platform. At the same time, by combining the information platform user browsing behavior trajectory data obtained by analysis (including the user's browsing path, click behavior, etc. on the information platform) to identify the browsing conversion process of the user, that is, the process of the user transferring from one page or content to another page or content, and by using changes The trend analysis method analyzes the browsing conversion process data of information platform users to analyze and understand the browsing conversion of users in different time periods and find out the trend of gradually decreasing conversion behavior of users in different time periods, that is, the trend of user interest decay. Secondly, the corresponding user browsing conversion process is analyzed using the impact assessment algorithm by combining the user browsing decreasing trend obtained by analysis to evaluate the decay of the user browsing conversion process and statistically analyze the decay law of the user browsing conversion process. Then, the user browsing behavior trajectory of the information platform obtained by combination analysis and the user browsing conversion decay law are used to mine and analyze the user browsing behavior in the corresponding information platform user browsing record data to find out the user's typical behavior patterns and preferences, and analyze the key nodes and conversion paths in the user's browsing conversion process, while exploring the user's behavior characteristics and behavior laws, and finally obtaining the information platform user browsing behavior data.

[0042] Step S13: Perform user search behavior mining and analysis on the information platform behavior log record data to obtain information platform user search behavior data;

[0043] The embodiment of the present invention first filters out records related to search behavior from the information platform behavior log record data, including search keywords, search result clicks, etc., and analyzes the user's search behavior characteristics, such as search frequency, popularity of search keywords, etc., by using a feature analysis method. At the same time, according to the search behavior characteristics obtained by the analysis, the information platform behavior log record data is mined and analyzed using a corresponding behavior analysis method to understand the user's search behavior on the information platform, including search keywords, search frequency, search result clicks, etc., and finally obtain the information platform user search behavior data.

[0044] Step S14: construct a user behavior preference matrix based on the information platform user browsing behavior data and the information platform user search behavior data to obtain an information platform user behavior preference feature matrix.

[0045] The embodiment of the present invention uses a feature analysis method to extract and analyze the browsing behavior data of information platform users and the search behavior data of information platform users respectively, so as to extract the feature information of user behavior, such as browsing behavior features such as browsing access frequency, browsing duration, browsing habits, browsing time period, and search behavior features such as search keywords, search frequency, and search preferences, and analyzes the browsing behavior features and search behavior features obtained by analysis by using a behavior preference index evaluation and analysis method, so as to comprehensively analyze the user's behavior feature data to establish a user behavior preference index, including information on the user's interest areas, content of concern, common functions, etc., and convert each user's behavior preference index into a feature At the same time, the user behavior preference indicator feature vectors obtained by analysis are classified by using a clustering algorithm or a classification algorithm, so as to classify the behavior preference indicator feature vectors of similar behavior patterns into the same category, and the classification results are annotated with classification labels to indicate the behavior preference category to which the feature vector belongs. Then, a matrix is ​​constructed by integrating the various behavior preference indicator feature vectors and their corresponding behavior preference indicator classification labels to integrate and combine the user's behavior preference feature vectors according to the classification labels, where each row represents a user, and each column represents the user's behavior preference feature vector and its corresponding classification label, and finally the information platform user behavior preference feature matrix is ​​obtained.

[0046] The present invention first obtains various behavior data of users on the information platform, such as browsing, searching, clicking, etc., by obtaining the behavior log record data of the information platform. These data record the activity trajectory and behavior pattern of users on the platform. Obtaining these data is crucial for understanding user behavior, analyzing user preferences, and optimizing platform services. By analyzing these data, users' interests and hobbies, behavioral habits can be gained, thereby providing data support for personalized recommendations and precision marketing of the information platform. Secondly, by mining and analyzing the user browsing behavior of the information platform behavior log record data, users' browsing behavior on the information platform can be deeply understood, including the pages visited, the dwell time, the clicked links, and other information. Such an analysis and processing process can help discover the user's interests and preferred areas, and then provide users with more personalized content recommendations and services that meet their needs. Then, by mining and analyzing the user search behavior of the information platform behavior log record data, users' search behavior on the information platform can be understood, including search keywords, search frequency, search result clicks, etc. Such an analysis and processing process helps to understand the user's needs and intentions, thereby providing users with more accurate and effective search suggestions and results display. Finally, by constructing a user behavior preference matrix based on the browsing behavior data of information platform users and the search behavior data of information platform users, the browsing behavior data of information platform users and the search behavior data of information platform users can be combined to construct a user behavior preference matrix. Through such matrix construction, the behavior preferences and interest characteristics of users on the information platform can be fully reflected, thereby forming a user behavior preference feature matrix. Such a feature matrix can provide an important data foundation and support for subsequent personalized recommendations, precision marketing and other businesses, thereby improving the service quality and user experience of the information platform.

[0047] Preferably, step S12 comprises the following steps:

[0048] Step S121: Perform user browsing record screening on the information platform behavior log record data to obtain information platform user browsing record data;

[0049] Step S122: Performing user browsing simulation restoration on the information platform user browsing record data to obtain information platform user browsing simulation restoration data;

[0050] Step S123: analyzing the browsing behavior trajectory of the information platform user's browsing simulation restoration data to obtain the information platform user's browsing behavior trajectory data;

[0051] Step S124: performing browsing conversion decay analysis on the browsing simulation restoration data of the information platform users based on the browsing behavior trajectory data of the information platform users, and obtaining browsing conversion decay law data of the information platform users;

[0052] Step S125: Based on the information platform user browsing behavior trajectory data and the information platform user browsing conversion decay law data, the information platform user browsing record data is mined and analyzed to obtain the information platform user browsing behavior data.

[0053] As an embodiment of the present invention, refer to Figure 3 As shown, Figure 2 Detailed step flow diagram of step S12 in the embodiment, step S12 includes the following steps:

[0054] Step S121: Perform user browsing record screening on the information platform behavior log record data to obtain information platform user browsing record data;

[0055] The embodiment of the present invention first collects the behavior log record data of the information platform, including various operation behaviors of users on the platform, such as browsing, clicking, searching, etc., and filters out the behavior information records related to user browsing according to pre-set filtering conditions, including visited pages, browsed content, etc. At the same time, the filtered data is cleaned and processed by removing invalid records, cleaning data, deduplicating and other operations to ensure the accuracy and completeness of the obtained user browsing record data, and finally obtain the information platform user browsing record data.

[0056] Step S122: Performing user browsing simulation restoration on the information platform user browsing record data to obtain information platform user browsing simulation restoration data;

[0057] The embodiment of the present invention analyzes and screens the browsing record data of information platform users to identify the browsing behavior sequence of users on the information platform, and simulates the user's browsing process according to the identified browsing behavior sequence to reproduce the user's browsing behavior on the information platform, including simulating user ID, browsing time, browsed pages or content and other information, and finally obtains information platform user browsing simulation restoration data.

[0058] Step S123: analyzing the browsing behavior trajectory of the information platform user's browsing simulation restoration data to obtain the information platform user's browsing behavior trajectory data;

[0059] The embodiment of the present invention uses a behavior trajectory analysis method to analyze the information platform user browsing simulation restoration data to analyze the user's browsing behavior trajectory, and understand the user's browsing habits, preferences, and information acquisition paths and methods on the information platform, and finally obtain the information platform user's browsing behavior trajectory data.

[0060] Step S124: performing browsing conversion decay analysis on the browsing simulation restoration data of the information platform users based on the browsing behavior trajectory data of the information platform users, and obtaining browsing conversion decay law data of the information platform users;

[0061] The embodiment of the present invention combines the information platform user browsing behavior trajectory data obtained by analysis (including the user's browsing path, click behavior, etc. on the information platform) to identify the browsing conversion process of the corresponding information platform user browsing simulation restoration data, so as to identify the user's browsing conversion process, that is, the process of the user transferring from one page or content to another page or content, and analyzes the information platform user browsing conversion process data by using a change trend analysis method to analyze and understand the user's browsing conversion in different time periods and find that the user's conversion behavior gradually decreases in different time periods, that is, the trend of user interest decay. At the same time, by combining the user browsing decreasing trend obtained by analysis, the corresponding user browsing conversion process is analyzed using an impact evaluation algorithm to evaluate the decay of the user browsing conversion process, and statistically analyze the factors affecting the conversion decay. Then, by combining the user browsing conversion decay influencing factors obtained by analysis, the corresponding information platform user browsing conversion process data is evolved using time series analysis, trend prediction and other methods to analyze the decay law of the user browsing conversion process, and understand the evolution trend and law of the conversion process, and finally obtain the information platform user browsing conversion decay law data.

[0062] Step S125: Based on the information platform user browsing behavior trajectory data and the information platform user browsing conversion decay law data, the information platform user browsing record data is mined and analyzed to obtain the information platform user browsing behavior data.

[0063] The embodiment of the present invention mines and analyzes the user browsing behavior in the corresponding information platform user browsing record data by combining the information platform user browsing behavior trajectory data and the information platform user browsing conversion decay law data obtained by analysis, so as to find out the user's typical behavior patterns and preferences, and analyze the key nodes and conversion paths in the user's browsing conversion process, while exploring the user's behavioral characteristics and behavioral laws, and finally obtaining the information platform user browsing behavior data.

[0064] The present invention firstly processes the user browsing records of the information platform behavior log record data to extract the relevant information data obtained by the user browsing. The screening process may include removing invalid records, cleaning data, deduplication and other operations to ensure the accuracy and completeness of the obtained user browsing record data. These processes will provide a reliable data basis for the subsequent user behavior analysis. Secondly, by performing user browsing simulation restoration on the information platform user browsing record data, that is, simulating the user's browsing path and behavior trajectory on the information platform according to the user's browsing behavior data, through such simulation restoration, the user's actual behavior can be restored more realistically, thereby providing a more accurate data basis for the subsequent behavior analysis. Then, by analyzing the browsing behavior trajectory of the information platform user browsing simulation restoration data, the browsing habits, preferences, and information acquisition paths and methods of the user on the information platform can be understood. Such an analysis and processing process helps to better understand user needs, thereby providing personalized services and recommendations for users. Next, by analyzing the browsing conversion decay analysis of the information platform user browsing simulation restoration data based on the information platform user browsing behavior trajectory data, in order to analyze the user browsing conversion decay law, we can understand the user's jump and conversion behavior during the browsing process, as well as the conversion rate and decay between different pages. Such an analysis and processing process helps to optimize the content structure and page design of the information platform and improve user browsing efficiency and satisfaction. Finally, by conducting user browsing behavior mining and analysis on the information platform user browsing record data based on the information platform user browsing behavior trajectory data and the information platform user browsing conversion decay law data, in order to analyze the user's browsing behavior, this step can deeply understand the user's interests, preferences, and reactions and choices to different content. Such analysis helps to provide users with more personalized and accurate content recommendations and services, and improve user experience and platform value.

[0065] Preferably, step S124 includes the following steps:

[0066] Based on the browsing behavior trajectory data of the information platform users, the browsing conversion process analysis is performed on the browsing simulation restoration data of the information platform users to obtain the browsing conversion process data of the information platform users;

[0067] The embodiment of the present invention combines the analyzed information platform user browsing behavior trajectory data (including the user's browsing path, click behavior, etc. on the information platform) with the corresponding information platform user browsing simulation restoration data to perform identification and analysis of the browsing conversion process, so as to identify the user's browsing conversion process, that is, the process of the user transferring from one page or content to another page or content, including information such as the page or content before the conversion, the page or content after the conversion, and the time of the conversion, and finally obtain the information platform user's browsing conversion process data.

[0068] Preferably, the information platform user browsing conversion process data is analyzed for conversion change trend to obtain the information platform user browsing conversion change trend data; the information platform user browsing conversion change trend data is analyzed for decreasing trend identification to obtain the information platform user browsing decreasing trend data;

[0069] The embodiment of the present invention analyzes the browsing conversion process data of information platform users by using a change trend analysis method to analyze and understand the browsing conversion situation of users in different time periods and find the change trend of users' conversion behavior in different time periods, thereby obtaining the browsing conversion change trend data of information platform users. Then, by performing a decreasing trend identification analysis on the browsing conversion change trend data of information platform users, the trend of gradually decreasing conversion behavior, that is, the trend of declining user interest, is identified, and finally the browsing decreasing trend data of information platform users is obtained.

[0070] Preferably, a conversion probability statistical analysis is performed on the information platform user browsing conversion process data to obtain the information platform user browsing conversion probability data;

[0071] The embodiment of the present invention uses a mathematical statistical analysis method to perform statistical analysis on the browsing conversion process data of information platform users to count the number of occurrences of each conversion process, and uses a probabilistic statistical model or algorithm (such as Markov chain, etc.) to calculate the number of occurrences of the conversion process obtained by analysis to quantify the probability of users converting from one page to another, and finally obtain the browsing conversion probability data of information platform users.

[0072] Preferably, based on the information platform user browsing decreasing trend data and the information platform user browsing conversion probability data, the information platform user browsing conversion process data is evaluated and analyzed for conversion decay impact, to obtain the information platform user browsing conversion decay influencing factor data;

[0073] The embodiment of the present invention uses an impact assessment algorithm to analyze the corresponding information platform user browsing conversion process data by combining the information platform user browsing decreasing trend data and the information platform user browsing conversion probability data obtained by analysis, so as to evaluate the decay of the user browsing conversion process, and statistically analyze the factors affecting the conversion decay, such as the reduction in conversion frequency, the change in conversion probability, etc., and finally obtain the information platform user browsing conversion decay influencing factor data.

[0074] Preferably, the decay law evolution analysis of the information platform user browsing conversion process data is performed based on the information platform user browsing conversion decay influencing factor data to obtain the information platform user browsing conversion decay law data.

[0075] The embodiment of the present invention uses the information platform user browsing conversion decay influencing factor data obtained by combining the analysis to perform evolutionary analysis on the corresponding information platform user browsing conversion process data using time series analysis, trend prediction and other methods, so as to analyze the decay law of the user browsing conversion process, understand the evolution trend and law of the conversion process, and finally obtain the information platform user browsing conversion decay law data.

[0076] The present invention first analyzes the browsing conversion process of the information platform user browsing simulation restoration data based on the information platform user browsing behavior trajectory data to analyze the user's browsing conversion process, including the user's conversion path, frequency, time and other information from one page to another. This step can deeply understand the user's browsing behavior pattern and conversion law on the information platform. Such an analysis and processing process helps to understand the user's behavioral motivation and needs, so as to provide guidance for optimizing the content structure and layout of the information platform. Secondly, by performing a conversion change trend analysis on the information platform user browsing conversion process data to analyze the changing trend of the user's conversion behavior in different time periods, the user's preference and trend change for conversion between different pages can be understood. At the same time, by performing a decreasing trend identification analysis on the information platform user browsing conversion change trend data, the trend of gradually decreasing conversion behavior, that is, the trend of user interest decay, can be identified, thereby providing data support for subsequent behavioral impact evaluation. Then, by conducting a conversion probability statistical analysis on the information platform user browsing conversion process data to analyze the conversion probability of users from one page to another, this step can understand the user's conversion preference and possibility between different pages, as well as the frequency and duration of conversion. Such analysis helps to better understand the user's behavior pattern, thereby providing a reference for personalized recommendation and content optimization. Next, by conducting a conversion decay impact evaluation analysis on the information platform user browsing conversion process data based on the information platform user browsing decreasing trend data and the information platform user browsing conversion probability data, in order to analyze the decay influencing factors in the conversion process, such as the reduction of conversion frequency, the change of conversion probability, etc., this step can evaluate the decay of the user browsing conversion process, thereby providing a data basis for the subsequent decay law evolution analysis. Finally, by conducting a decay law evolution analysis on the information platform user browsing conversion process data based on the information platform user browsing conversion decay influencing factor data to analyze the evolution trend of the decay law, the decay speed and mode of user interest, as well as the factors and mechanisms that affect the decay of user interest, such analysis helps to better understand the dynamic changes of user behavior, thereby providing strategic suggestions for content recommendation and service optimization of the information platform.

[0077] Preferably, step S14 comprises the following steps:

[0078] Step S141: performing behavior characteristic analysis on the information platform user browsing behavior data and the information platform user search behavior data to obtain the information platform user browsing behavior characteristic data and the information platform user search behavior characteristic data;

[0079] The embodiments of the present invention use a feature analysis method to extract and analyze the browsing behavior data of information platform users and the search behavior data of information platform users, so as to extract feature information of user behavior, such as browsing behavior features such as browsing access frequency, browsing duration, browsing habits, and browsing time periods, as well as search behavior features such as search keywords, search frequency, and search preferences, and finally obtain the browsing behavior feature data of information platform users and the search behavior feature data of information platform users.

[0080] Step S142: performing behavior preference index evaluation and analysis based on the information platform user browsing behavior characteristic data and the information platform user search behavior characteristic data to obtain information platform user behavior preference index data;

[0081] The embodiment of the present invention uses a behavior preference index evaluation and analysis method to analyze the browsing behavior characteristic data of information platform users and the search behavior characteristic data of information platform users, so as to comprehensively analyze the user's behavior characteristic data to establish a user behavior preference index for describing and measuring the user's preference and degree of preference for the content of the information platform, including information on the user's interest areas, content of concern, common functions, etc., and finally obtain the information platform user behavior preference index data.

[0082] Step S143: Performing feature vector conversion analysis on the information platform user behavior preference index data to obtain the information platform user behavior preference index feature vector;

[0083] The embodiment of the present invention uses a feature engineering method to perform feature vectorization conversion processing on the information platform user behavior preference index data obtained through evaluation and analysis, so as to convert the behavior preference index of each user into the form of a feature vector, which includes the numerical representation of the user on different behavior preference indicators, and finally obtains the information platform user behavior preference index feature vector.

[0084] Step S144: performing user classification labeling processing on the information platform user behavior preference index feature vector to obtain the information platform user behavior preference index classification label;

[0085] The embodiment of the present invention classifies the information platform user behavior preference indicator feature vectors obtained by analysis by using a clustering algorithm or a classification algorithm to divide the behavior preference indicator feature vectors of similar behavior patterns into the same category, and annotates the classification results with classification labels to indicate the behavior preference category to which the feature vector belongs, and finally obtains the information platform user behavior preference indicator classification label.

[0086] Step S145: construct a user behavior preference matrix according to the information platform user behavior preference indicator feature vector and the information platform user behavior preference indicator classification label to obtain the information platform user behavior preference feature matrix.

[0087] The embodiment of the present invention constructs a matrix by integrating the characteristic vectors of user behavior preference indicators of each information platform and their corresponding classification labels of user behavior preference indicators of the information platform, so as to integrate and combine the user behavior preference characteristic vectors according to the classification labels, wherein each row represents a user, and each column represents the user's behavior preference characteristic vector and its corresponding classification label, and finally an information platform user behavior preference characteristic matrix is ​​obtained.

[0088] The present invention firstly performs behavioral feature analysis on the browsing behavior data of information platform users and the search behavior data of information platform users to extract the characteristic information of user behavior. By analyzing the browsing and search behaviors of users, such as access frequency, browsing time, search keywords, etc., the behavioral feature data of users on the information platform can be obtained. These feature data include information on the user's interest preferences, browsing habits, search preferences, etc., thereby providing a data basis for the subsequent evaluation of behavioral preference indicators. Secondly, by performing behavioral preference index evaluation and analysis based on the browsing behavior feature data of information platform users and the search behavior feature data of information platform users, the user behavior preference index is established by comprehensively analyzing the user's behavioral feature data to describe and measure the user's preference and preference for the content of the information platform. These behavioral preference indicators include information on the user's interest areas, content of concern, common functions, etc., thereby providing data support for the subsequent construction of the user behavior preference matrix. Then, by performing feature vector conversion analysis on the information platform user behavior preference index data, the user behavior preference index data can be converted into feature vectors. This step can quantify the user's behavior characteristics into numerical form, which is convenient for subsequent data processing and analysis. These feature vectors contain the numerical representation of the user on different behavior preference indicators, thus laying the foundation for the construction of the user behavior preference matrix. Next, by performing user classification labeling processing on the information platform user behavior preference index feature vector, users can be divided into different behavior preference categories. These classification labels reflect the similarities and differences of users in behavior preference indicators, thus providing a classification basis for the construction of the user behavior preference matrix. Finally, by constructing the user behavior preference matrix according to the information platform user behavior preference index feature vector and the information platform user behavior preference index classification label, the user behavior preference feature matrix is ​​constructed. By integrating and combining the user behavior preference feature vector according to the classification label, the user behavior preference feature matrix can be obtained. This feature matrix reflects the user's behavior characteristics and preference degree under different behavior preference categories, thus providing a data basis for personalized recommendation and content customization.

[0089] Preferably, step S2 comprises the following steps:

[0090] Step S21: Perform browsing information screening on the information platform behavior log record data to obtain a user browsing related information data set;

[0091] The embodiment of the present invention collects behavioral log record data of an information platform, including users' browsing, clicking, searching and other behaviors, and screens the browsing information of the information platform behavior log record data to clean, filter and classify the information platform behavior log record data, and removes irrelevant information to extract information data sets related to user browsing behaviors, such as browsed web pages, read articles and other information, and finally obtains user browsing related information data sets.

[0092] Step S22: performing context association mining analysis on the user browsing related information data set to obtain context information association rule data between each user browsing information data;

[0093] The embodiment of the present invention processes the information data browsed by each user in the user browsing related information data set by using text mining technology (such as natural language processing tools, keyword extraction algorithms, etc.) to identify and extract the most representative and key text information in the user browsing information, and the key text information includes title, abstract, keywords, etc., and evaluates and analyzes the key text information content data obtained by mining analysis by using text similarity calculation methods (such as cosine similarity, Jaccard similarity, etc.) to compare the text information content between different user browsing information data to quantitatively calculate the relevance of the text information content between different user browsing information data, and secondly, by combining the text information content obtained by analysis, the text information content is compared with the text information content of different users. The degree of relevance uses a graph theory algorithm (such as a minimum spanning tree algorithm and a shortest path algorithm) to perform a matching analysis on the contextual connection relationship between the corresponding user browsing information data in the user browsing related information data set, so as to determine the contextual association matching relationship between the various user browsing information data; then, based on the contextual association matching relationship obtained by the analysis, an association rule mining algorithm (such as an Apriori algorithm and an FP-Growth algorithm) is used to mine and analyze the association rules between the various user browsing information data in the user browsing related information data set, so as to discover the contextual information association laws and patterns between the user browsing information data, and finally obtain the contextual information association rule data between the various user browsing information data.

[0094] Step S23: performing context change trend analysis on the user browsing related information data set based on the context information association rule data between each user browsing information data, and obtaining context information association change trend data between each user browsing information data;

[0095] The embodiment of the present invention infers the changing trend of the context information association of the corresponding user browsing information data in the user browsing related information data set by combining the context information association rule data between the various user browsing information data obtained by analysis, and reveals the changing trend of the association between the user browsing information data, including the increase or decrease of the degree of association between the user browsing information data, the rate of change, the context information associated with the user browsing information data, etc., and finally obtains the context information association changing trend data between the various user browsing information data.

[0096] Step S24: sorting the user browsing related information data sets according to the context information association change trend data between each user browsing information data, and obtaining the information context association perception sequence data set.

[0097] The embodiment of the present invention uses a corresponding sorting algorithm (such as sorting based on frequent item sets, sorting based on confidence, etc.) to sort the corresponding user browsing information data in the user browsing related information data set according to the context information connection relationship by combining the context information association change trend data between the various user browsing information data obtained by analysis, and finally obtains the information context association perception sequence data set.

[0098] The present invention firstly processes browsing information screening of information platform behavior log record data to extract user browsing related information data sets. This step can remove irrelevant information by cleaning, filtering and classifying behavior log data to extract user browsing behavior related data sets. This data set contains the user's browsing records on the information platform, thereby providing a data basis for subsequent context association mining and analysis. Secondly, by performing context association mining and analysis on user browsing related information data sets to mine and analyze the association rules between user browsing information data, this step can discover the context association of users when browsing information. These association rules reflect the degree of association and occurrence pattern between different information, thereby providing in-depth insights into user browsing behavior. Then, by performing context change trend analysis on user browsing related information data sets based on the context information association rule data between each user browsing information data, in order to analyze the context association change trend between user browsing information data, the user's browsing preferences and change trends in different time periods or different situations can be understood. These trend data are helpful to grasp the user's interest changes and behavior trends. Next, by sorting the association perception response of the user browsing related information data set according to the context information association change trend data between each user's browsing information data, we can obtain the information context association perception sequence data set, in which each information is arranged according to the connection relationship of its corresponding context information. This sequence data set reflects that the information data in the user browsing related information data set is sorted according to the context information connection relationship, thereby providing an important basis for personalized recommendation and content push.

[0099] Preferably, step S22 includes the following steps:

[0100] Perform key text information mining and analysis on the user browsing related information data set to obtain the key text information content data corresponding to each user browsing information data;

[0101] The embodiment of the present invention processes the information data browsed by each user in the user browsing related information data set by using text mining technology (such as natural language processing tools, keyword extraction algorithms, etc.) to identify and extract the most representative and key text information in the user browsing information, and these key text information include titles, abstracts, keywords, etc., and finally obtain the key text information content data corresponding to each user browsing information data.

[0102] Preferably, a correlation evaluation analysis is performed on the key text information content data corresponding to each user browsing information data to obtain the text information content correlation degree data between each user browsing information data;

[0103] The embodiment of the present invention uses a text similarity calculation method (such as cosine similarity, Jaccard similarity, etc.) to evaluate and analyze the key text information content data corresponding to each user browsing information data, so as to compare the text information content between different user browsing information data to quantitatively calculate the text information content correlation between different user browsing information data, and finally obtain the text information content correlation data between each user browsing information data.

[0104] Preferably, context association matching analysis is performed on each user browsing information data in the user browsing related information data set based on the text information content correlation data between each user browsing information data, so as to obtain the context association matching connection relationship between each user browsing information data;

[0105] The embodiment of the present invention uses a graph theory algorithm (such as a minimum spanning tree algorithm and a shortest path algorithm) to match and analyze the contextual connection relationship between the corresponding user browsing information data in the user browsing related information data set by combining and analyzing the text information content correlation data between the various user browsing information data, so as to determine the contextual correlation matching relationship between the various user browsing information data. This connection relationship reflects the correlation degree and mutual influence between the user browsing behaviors, and finally obtains the contextual correlation matching connection relationship between the various user browsing information data.

[0106] Preferably, association rule mining analysis is performed on each user browsing information data in the user browsing related information data set according to the context association matching connection relationship between each user browsing information data, so as to obtain context information association rule data between each user browsing information data.

[0107] The embodiment of the present invention uses an association rule mining algorithm (such as an Apriori algorithm and an FP-Growth algorithm) to mine and analyze the association rules between each user browsing information data in a user browsing related information data set by combining the context association matching connection relationship between each user browsing information data obtained by analysis, so as to discover the context information association rules and patterns between the user browsing information data, and finally obtain the context information association rule data between each user browsing information data.

[0108] The present invention firstly performs key text information mining and analysis on the user browsing related information data set to extract the key text content of each user browsing information. This step can identify and extract the most representative and key text information in the user browsing information by using natural language processing technology and text analysis algorithm. These key text information include titles, abstracts, keywords, etc., thereby providing a data basis for subsequent correlation evaluation and context association analysis. Secondly, by performing correlation evaluation and analysis on the key text information content data corresponding to each user browsing information data, it is possible to compare the text information content between different user browsing information data and calculate the correlation degree between them. This correlation evaluation can be performed based on indicators such as text similarity and semantic association, thereby quantifying the text information content correlation degree between different user browsing information data, and providing data support for subsequent context association matching. Then, by performing context association matching analysis on each user browsing information data in the user browsing related information data set based on the text information content correlation degree data between each user browsing information data, it is possible to find the context association relationship between different user browsing information data, thereby establishing a matching connection relationship between them, and this connection relationship reflects the correlation degree and mutual influence between user browsing behaviors, thereby providing a basis for subsequent association rule mining. Finally, by performing association rule mining and analysis on each user browsing information data set based on the contextual association matching connection relationship between each user browsing information data, this step can discover the contextual information association rules and patterns between user browsing information data. These association rules can help understand user browsing behavior and discover potential associations between information, thereby providing support for personalized recommendations and content matching.

[0109] Preferably, step S23 includes the following steps:

[0110] Step S231: performing association rule evolution analysis on the user browsing related information data set based on the context information association rule data between each user browsing information data, and obtaining context information association evolution process data between each user browsing information data;

[0111] The embodiment of the present invention performs evolutionary analysis on the association rules between the corresponding user browsing information data in the user browsing related information data set by combining the context information association rule data between the various user browsing information data obtained through analysis, so as to track the changes in the association rules between the user browsing information data, and reveal the evolution process of these association rules with time or other changes, and finally obtain the context information association evolution process data between the various user browsing information data.

[0112] Step S232: plotting an evolution curve of the context information association evolution process data between each user browsing information data, so as to obtain a context information association evolution mapping curve between each user browsing information data;

[0113] The embodiment of the present invention organizes and compiles the context information association evolution process data between each user's browsing information data, and draws an evolution curve by using a visualization tool (such as Matplotlib or Plotly) to map the context information association evolution between each user into a curve graph, and intuitively show the association evolution trend between the user's browsing information data, and finally obtains the context information association evolution mapping curve between the browsing information data of each user.

[0114] Step S233: performing an evolution series error approximation analysis on the context information association evolution mapping curve between each user browsing information data, so as to obtain context information evolution series error approximation data between each user browsing information data;

[0115] The embodiment of the present invention first uses a mathematical method (such as the least squares method) to perform an error analysis on the context information association evolution mapping curves between each user's browsing information data to calculate the evolution error between the context information association evolution mapping curves, and then uses the Chebyshev series to perform series approximation statistical analysis on the evolution error between the calculated context information association evolution mapping curves to quantitatively evaluate the approximation degree of the association evolution mapping curve, and finally obtains the context information evolution series error approximation data between each user's browsing information data.

[0116] Step S234: performing error fitting analysis on the context information association evolution mapping curve between each user browsing information data according to the context information evolution series error approximation data between each user browsing information data, and obtaining the context information evolution error fitting curve between each user browsing information data;

[0117] The embodiment of the present invention performs error fitting analysis on the corresponding context information association evolution mapping curve between each user browsing information data by combining the context information evolution series error approximation data between each user browsing information data obtained through analysis, so as to find the best fitting curve according to the corresponding evolution series error approximation data, thereby better describing the association evolution trend between the user browsing information data, and finally obtaining the context information evolution error fitting curve between each user browsing information data.

[0118] Step S235: performing context change trend analysis on the context information evolution error fitting curve between each user browsing information data, and obtaining context information association change trend data between each user browsing information data.

[0119] The embodiment of the present invention infers the changing trend of the context information association of the corresponding user browsing information data by analyzing the slope, curvature and other characteristics of the context information evolution error fitting curve between each user browsing information data, and reveals the changing trend of the association between the user browsing information data, including the increase or decrease in the degree of association between the user browsing information data, the rate of change, the context information associated with the user browsing information data, etc., and finally obtains the context information association changing trend data between each user browsing information data.

[0120] The present invention firstly performs association rule evolution analysis on user browsing related information data sets based on context information association rule data between each user browsing information data to track the association rule changes between user browsing information data. This step can reveal the process of these rules evolving over time or other changes. These evolution process data provide important insights into changes in user behavior and preferences, which are helpful to understand the changing trend of user interests and the evolution of information browsing patterns. Secondly, by drawing evolution curves for the context information association evolution process data between each user browsing information data, this step can intuitively show the association evolution trend between user browsing information data by drawing evolution curves. These curves reflect the changing trend of user behavior and preferences, thereby providing a visualization tool for understanding the dynamic changes of user information browsing behavior. Then, by performing evolution series error approximation analysis on the context information association evolution mapping curve between each user browsing information data, the changing trend and pattern of the evolution curve can be analyzed, and the evolution series error can be quantified, thereby evaluating the approximation degree of the association evolution mapping curve. These error approximation data can help understand the accuracy and stability of the evolution curve, thereby providing a basis for subsequent error fitting analysis. Next, by performing error fitting analysis on the context information association evolution mapping curve between each user browsing information data according to the context information evolution series error approximation data between each user browsing information data, the best fitting curve can be found according to the corresponding evolution series error approximation data, so as to better describe the association evolution trend between user browsing information data. These fitting curves provide a more accurate description of user behavior changes, thus providing a more in-depth analysis tool for understanding user interests and behavior patterns. Finally, by performing context change trend analysis on the context information evolution error fitting curve between each user browsing information data to analyze the change trend of the fitting curve, this step can reveal the association change trend between user browsing information data. These change trend data provide an in-depth understanding of user behavior and interest changes, which helps to adjust information recommendation strategies and optimize user experience.

[0121] Preferably, step S3 comprises the following steps:

[0122] Step S31: performing context information perception configuration on the information context association perception sequence data set to obtain information context perception configuration result data;

[0123] The embodiment of the present invention determines the goals and requirements of information context awareness configuration through the information context association connection relationship of the information context association awareness sequence data set, including the required context information and configuration parameters, and connects and configures the context information of the corresponding information content in the information context association awareness sequence data set according to the determined goals and requirements, so as to identify and understand the user's behavior and preferences on the information platform, and ensure that the context of the information can be perceived and understood, and finally obtain the information context awareness configuration result data.

[0124] Step S32: performing information push rule mining and analysis based on the information platform user behavior preference feature matrix and the information context perception configuration result data to obtain the information platform user preference information push initial rules;

[0125] The embodiment of the present invention mines and analyzes the push rules of corresponding information data based on the behavior preference feature parameters (including click-through rate, browsing history, subscription preference, etc.) in the information platform user behavior preference feature matrix and the connected and configured information context perception configuration result data, so as to analyze the association between user behavior preferences and context information, and discover the user's preference patterns and behavior regularities, and then determine the initial information push rules, including the time, frequency and content of information push, and finally obtain the initial rules for information platform user preference information push.

[0126] Step S33: performing information push potential impact assessment analysis on the information platform user behavior preference feature matrix and the information context perception configuration result data to obtain information push rule potential impact pattern data between user behavior preference and information context association data;

[0127] The embodiment of the present invention uses an impact assessment algorithm to analyze the behavior preference feature parameters (including click-through rate, browsing history, subscription preference, etc.) in the user behavior preference feature matrix of the information platform and the information context-aware configuration result data to evaluate the potential impact relationship between them, and explore the potential impact pattern between user behavior preferences and information context-related data to understand the relationship pattern and impact degree between them, and finally obtain the potential impact pattern data of information push rules between user behavior preferences and information context-related data.

[0128] Step S34: optimizing the initial information push rules of the information platform user preference based on the potential impact pattern data of the information push rules between the user behavior preference and the information context association data, so as to obtain the information platform information push rule data set;

[0129] The embodiment of the present invention optimizes the mined initial rules for information push of user preferences of the information platform by using the potential impact pattern data of information push rules between user behavior preferences and information context association data obtained by evaluation and analysis, so as to deeply explore the deep association relationship between user behavior data and context information to optimize the corresponding information push rules (including the time, frequency and content of information push, etc.), and improve the accuracy and personalization of information push, and finally obtain the information platform information push rule data set.

[0130] Step S35: construct a customized window according to the information platform information push rule data set to generate an information platform information push customized window.

[0131] The embodiment of the present invention designs the parameters and functions of the customized window by combining the generated information platform information push rule data set, including push content, push frequency, push time, etc., to ensure that it can intelligently push information according to the user's behavioral preferences and contextual information content connection relationship, and present the most relevant and attractive information to the user, and finally generate an information platform information push customized window.

[0132] The present invention firstly configures the context information perception of the information context association perception sequence data set, and can accurately identify and understand the user's behavior and preference on the information platform. By analyzing the user's context information, such as browsing history, search history, etc., the user's context information model can be established, so as to better grasp the user's interests and needs, and provide basic data for subsequent information push. Secondly, by mining and analyzing the information push rule according to the information platform user behavior preference feature matrix and the information context perception configuration result data, the user's preference pattern and behavior law can be found to analyze the user's behavior preference and context information, and then the initial information push rule can be determined, which can help to more accurately push the information of interest to the user, thereby improving the accuracy and effect of information push. Then, by evaluating and analyzing the potential impact of information push on the information platform user behavior preference feature matrix and the information context perception configuration result data, the potential impact pattern between the user behavior preference and the information context association data can be deeply excavated, and the law and trend hidden behind the data can be found by analyzing the correlation between the user behavior and the context information, and the information push rule can be further optimized, thereby improving the accuracy and personalization of information push. Next, the initial information push rules of the information platform user preference are optimized based on the potential impact pattern data of the information push rules between the user behavior preference and the information context association data, so as to deeply explore the association between the user behavior data and the context information. This step can optimize the information push rules more accurately, thereby improving the effect of information push and user satisfaction. The optimized information push rule data set can provide better basic support for subsequent information push. Finally, a customized window is constructed according to the information platform information push rule data set to generate a personalized information push interface. Through the information push rules generated according to the user behavior preference and context information, the content and form of the information push can be dynamically adjusted to present the most relevant and attractive information to users. The construction of this step of customized window can provide users with a better information browsing experience, thereby enhancing the user's stickiness and loyalty to the information platform.

[0133] Preferably, step S4 comprises the following steps:

[0134] Step S41: Obtaining the information data set to be pushed;

[0135] The embodiment of the present invention obtains information data sets to be recommended through information platforms or other data sources, including but not limited to various information contents such as articles, news, and advertisements, and finally obtains information data sets to be pushed.

[0136] Step S42: performing statistical calculation of the matching degree of the information data set to be pushed according to the information platform information push customization window, so as to obtain the matching degree between the historical pushed information in the customization window and the information to be pushed;

[0137] The embodiment of the present invention uses a previously designed information platform information push customized window and uses a matching algorithm to perform matching calculations on the information to be pushed in the information data set and the historical pushed information in the customized window, so as to evaluate the degree of matching between the information to be pushed and the historical push information of user interests in the customized window, and finally obtain the degree of matching between the historical push information in the customized window and the information to be pushed.

[0138] Step S43: performing matching sorting processing on the information data set to be pushed based on the matching degree between the historical pushed information and the information to be pushed in the customized window, and obtaining a sequence of matching information to be pushed;

[0139] The embodiment of the present invention sorts the information to be pushed corresponding to the information to be pushed data set according to the degree of matching between the historical pushed information and the information to be pushed in the quantified generated customized window, so as to put the information with a higher matching degree in front, and determines the priority and order of the matching information to be pushed, thereby ensuring that the user receives the most relevant and valuable information first, and finally obtains a sequence of matching information to be pushed.

[0140] Step S44: Intelligently and regularly push the matching information sequence to be pushed through the information platform information push customization window to obtain an information intelligent matching push sequence set.

[0141] The embodiment of the present invention uses the information platform information push customization window to perform timed push processing on the information to be pushed sorted in the push matching information sequence according to the corresponding push frequency, push time and other push rules, so as to ensure that the user can receive the relevant interest information at the appropriate time, and ensure that the user receives the information push at the most appropriate time, and finally obtain the information intelligent matching push sequence set.

[0142] The present invention first obtains the information data set to be pushed, including but not limited to various information contents such as articles, news, and advertisements. This step can provide users with diversified and personalized information content by collecting the information data set to be pushed to meet their different needs and interests. Secondly, by performing statistical calculations on the matching degree of the information data set to be pushed according to the information platform information push customization window, the matching degree between the historical push information in the customization window and the information to be pushed can be quantitatively analyzed. This step can evaluate the matching degree between the information to be pushed and the user's interests, so as to decide whether to push the information to the user, so as to ensure that the information pushed to the user is closer to his interests and needs, and improve the accuracy and effect of information push. Then, by matching and sorting the information data set to be pushed based on the matching degree between the historical push information in the customization window and the information to be pushed, the information to be pushed can be sorted according to the matching degree with the user's interests, and the priority and order of the matching information to be pushed can be determined, so as to ensure that the user first receives the most relevant and valuable information, so as to improve the user's acceptance and satisfaction with information push. Finally, by using the generated information platform information push customization window, the matched information sequence to be pushed is intelligently timed and pushed to determine the best push timing based on factors such as the user's active time and preferences. This step can ensure that users receive information push at the most appropriate time, increase user attention to information and click-through rate, and the intelligent timed push processing process can also reduce user interference with unnecessary information, thereby improving the effect of information push and user experience.

[0143] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0144] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for intelligently pushing information data based on a customized window, characterized in that: The following steps are involved: Step S1: obtaining information platform behavior log record data, and performing user browsing behavior and search behavior mining and analysis on the information platform behavior log record data to obtain information platform user browsing behavior data and information platform user search behavior data; Constructing a user behavior preference matrix based on the browsing behavior data of information platform users and the search behavior data of information platform users to obtain an information platform user behavior preference feature matrix; Step S2: Perform browsing information screening on the information platform behavior log record data to obtain a user browsing related information data set; Perform context association perception analysis on the user browsing related information dataset to obtain the information context association perception sequence dataset; Step S3 includes: performing information push potential impact assessment analysis on the information platform user behavior preference feature matrix and the information context perception configuration result data to obtain the information push rule potential impact pattern data between the user behavior preference and the information context association data; performing push rule optimization processing on the information platform user preference information push initial rules based on the information behavior preference and the information context association data to obtain the information platform information push rule data set; performing customized window construction based on the information platform information push rule data set to generate the information platform information push customized window; Step S4: Acquire the information data set to be pushed, and perform intelligent matching and pushing processing on the information data set to be pushed according to the information platform information push customization window to obtain an information intelligent matching push sequence set.

2. The method for intelligently pushing information data based on customized windows according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Obtaining information platform behavior log record data; Step S12: performing user browsing behavior mining and analysis on the information platform behavior log record data to obtain information platform user browsing behavior data; Step S13: Perform user search behavior mining and analysis on the information platform behavior log record data to obtain information platform user search behavior data; Step S14: construct a user behavior preference matrix based on the information platform user browsing behavior data and the information platform user search behavior data to obtain an information platform user behavior preference feature matrix.

3. The method for intelligently pushing information data based on customized windows according to claim 2, characterized in that: Step S12 includes the following steps: Step S121: Perform user browsing record screening on the information platform behavior log record data to obtain information platform user browsing record data; Step S122: Performing user browsing simulation restoration on the information platform user browsing record data to obtain information platform user browsing simulation restoration data; Step S123: analyzing the browsing behavior trajectory of the information platform user's browsing simulation restoration data to obtain the information platform user's browsing behavior trajectory data; Step S124: performing browsing conversion decay analysis on the browsing simulation restoration data of the information platform users based on the browsing behavior trajectory data of the information platform users, and obtaining browsing conversion decay law data of the information platform users; Step S125: Based on the information platform user browsing behavior trajectory data and the information platform user browsing conversion decay law data, the information platform user browsing record data is mined and analyzed to obtain the information platform user browsing behavior data.

4. The method for intelligently pushing information data based on customized windows according to claim 3, characterized in that: Step S124 includes the following steps: Based on the browsing behavior trajectory data of the information platform users, the browsing conversion process analysis is performed on the browsing simulation restoration data of the information platform users to obtain the browsing conversion process data of the information platform users; Perform conversion change trend analysis on the browsing conversion process data of the information platform users to obtain the browsing conversion change trend data of the information platform users; perform decreasing trend identification analysis on the browsing conversion change trend data of the information platform users to obtain the browsing decreasing trend data of the information platform users; Conduct conversion probability statistical analysis on the browsing conversion process data of information platform users to obtain browsing conversion probability data of information platform users; Based on the decreasing trend data of information platform users’ browsing and the probability data of information platform users’ browsing conversion, the conversion decay impact evaluation analysis is conducted on the information platform users’ browsing conversion process data to obtain the data of factors influencing the information platform users’ browsing conversion decay; According to the decay influencing factors data of information platform user browsing conversion decay, the decay law evolution analysis of information platform user browsing conversion process data is carried out to obtain the information platform user browsing conversion decay law data.

5. The method for intelligently pushing information data based on customized windows according to claim 2, characterized in that: Step S14 includes the following steps: Step S141: performing behavior characteristic analysis on the information platform user browsing behavior data and the information platform user search behavior data to obtain the information platform user browsing behavior characteristic data and the information platform user search behavior characteristic data; Step S142: performing behavior preference index evaluation and analysis based on the information platform user browsing behavior characteristic data and the information platform user search behavior characteristic data to obtain information platform user behavior preference index data; Step S143: Performing feature vector conversion analysis on the information platform user behavior preference index data to obtain the information platform user behavior preference index feature vector; Step S144: performing user classification labeling processing on the information platform user behavior preference index feature vector to obtain the information platform user behavior preference index classification label; Step S145: construct a user behavior preference matrix based on the information platform user behavior preference indicator feature vector and the information platform user behavior preference indicator classification label to obtain the information platform user behavior preference feature matrix.

6. The method for intelligently pushing information data based on customized windows according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: Perform browsing information screening on the information platform behavior log record data to obtain a user browsing related information data set; Step S22: performing context association mining analysis on the user browsing related information data set to obtain context information association rule data between each user browsing information data; Step S23: performing context change trend analysis on the user browsing related information data set based on the context information association rule data between each user browsing information data, and obtaining context information association change trend data between each user browsing information data; Step S24: sorting the user browsing related information data sets according to the context information association change trend data between each user browsing information data, and obtaining the information context association perception sequence data set.

7. The method for intelligently pushing information data based on customized windows according to claim 6, characterized in that: Step S22 includes the following steps: Perform key text information mining and analysis on the user browsing related information data set to obtain the key text information content data corresponding to each user browsing information data; Performing correlation evaluation and analysis on key text information content data corresponding to each user's browsing information data to obtain text information content correlation data between each user's browsing information data; Based on the text information content relevance data between each user browsing information data, context association matching analysis is performed on each user browsing information data set in the user browsing related information data set to obtain the context association matching connection relationship between each user browsing information data; According to the context association matching connection relationship between each user browsing information data, each user browsing information data in the user browsing related information data set is subjected to association rule mining analysis to obtain context information association rule data between each user browsing information data.

8. The method for intelligently pushing information data based on customized windows according to claim 6, characterized in that: Step S23 includes the following steps: Step S231: performing association rule evolution analysis on the user browsing related information data set based on the context information association rule data between each user browsing information data, and obtaining context information association evolution process data between each user browsing information data; Step S232: plotting an evolution curve of the context information association evolution process data between each user browsing information data, so as to obtain a context information association evolution mapping curve between each user browsing information data; Step S233: performing an evolution series error approximation analysis on the context information association evolution mapping curve between each user browsing information data, so as to obtain context information evolution series error approximation data between each user browsing information data; Step S234: performing error fitting analysis on the context information association evolution mapping curve between each user browsing information data according to the context information evolution series error approximation data between each user browsing information data, and obtaining the context information evolution error fitting curve between each user browsing information data; Step S235: performing context change trend analysis on the context information evolution error fitting curve between each user browsing information data, and obtaining context information association change trend data between each user browsing information data.

9. The method for intelligently pushing information data based on customized windows according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: Obtaining the information data set to be pushed; Step S42: performing statistical calculation of the matching degree of the information data set to be pushed according to the information platform information push customization window, so as to obtain the matching degree between the historical pushed information in the customization window and the information to be pushed; Step S43: performing matching sorting processing on the information data set to be pushed based on the matching degree between the historical pushed information and the information to be pushed in the customized window, and obtaining a sequence of matching information to be pushed; Step S44: Intelligently and regularly push the matching information sequence to be pushed through the information platform information push customization window to obtain an information intelligent matching push sequence set.

Images