Methods, devices, electronic devices, and storage media for user behavior analysis and prediction.

Abstract

This application provides a method, apparatus, electronic device, and storage medium for user behavior analysis and prediction. The user behavior analysis includes: acquiring historical user behavior data; obtaining analysis results based on the historical user behavior data; and forming a behavior data benchmark library, a user benchmark library, and a data popularity library based on the analysis results. The user benchmark library includes users who share similar interests in data and business. This application can form the behavior data benchmark library, user benchmark library, and data popularity library, thereby fully utilizing historical user behavior data and enhancing its value. Simultaneously, this application can proactively prompt users for data they are interested in using the behavior data benchmark library, the user benchmark library, and the data popularity library, reducing user input search behavior, improving user experience in business scenarios, enhancing the value and efficiency of behavior data utilization, and ultimately improving user experience.

Description

Technical Field

[0001] This application relates to the field of big data, and more specifically, to a method, apparatus, electronic device, and storage medium for user behavior analysis and prediction. Background Technology

[0002] User behavior data refers to the action data generated by users during their interactions with the system, i.e., what the user did. For example, a user might query a security incident or retrieve a risk report. Most systems collect and record user behavior data as audit data without identifying or analyzing it.

[0003] In practical business operations, we often consider meeting user needs and providing better products and services as our goals. Beyond the explicit needs provided by users, can we uncover their latent needs through other means? The answer lies in deeply understanding and analyzing user needs using the collected and recorded user behavior data. We need to analyze user behavior to understand user needs, guide business upgrades, predict user behavior, and drive business innovation.

[0004] To address the aforementioned needs, existing technologies can only perform statistical analysis on historical user behavior, relying on historical data for user behavior visualization. However, they cannot analyze the continuity of historical user behavior, thus limiting their applicability. Secondly, existing technologies offer limited recording and analysis of user behavior, resulting in insufficient support for user-related business needs and inadequate utilization of data value. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, electronic device, and storage medium for user behavior analysis and prediction, in order to address the technical deficiencies of the prior art.

[0006] Firstly, this application provides a user behavior analysis method, the method comprising:

[0007] Obtain user historical behavior data;

[0008] The analysis results are obtained based on the user's historical behavior data, and a behavior data benchmark library, a user benchmark library, and a data popularity library are formed based on the analysis results. The user benchmark library includes users who have the same data and business interests.

[0009] Firstly, this application enables the acquisition of historical user behavior data, and subsequently, the generation of analytical results based on this data. Based on these results, a behavior data benchmark library, a user benchmark library, and a data popularity library are formed. The user benchmark library includes users with similar data and business interests. Compared to existing technologies, this application can form these three libraries, which can be used to predict users' continued behavior. This allows for full utilization of historical user behavior data and enhances its value. Existing technologies, however, can only statistically analyze user data, resulting in limited available data and an inability to analyze the continuity of users' historical behavior, thus limiting their applicability. Secondly, existing technologies offer relatively limited recording and analysis of user behavior, leading to insufficient support for user business and inadequate utilization of data value.

[0010] In an optional implementation, the step of obtaining analysis results based on the user's historical behavior data, and forming a behavior data benchmark library, a user benchmark library, and a data popularity library based on the analysis results, includes:

[0011] Identify the operation data and retrieval data in the user's historical behavior data, and use the operation data and retrieval data as raw behavior data, and associate the raw behavior data with the user ID;

[0012] The retrieved data is analyzed based on NLP (Natural Language Processing) algorithms to obtain the behavioral data benchmark library;

[0013] The user benchmark library is obtained based on the aforementioned behavioral data benchmark library and the Euclidean distance algorithm;

[0014] The data popularity database is calculated based on the user benchmark database and the behavioral data benchmark database.

[0015] This optional implementation can identify operation data and retrieval data in the user's historical behavior data, and use the operation data and retrieval data as raw behavior data. The raw behavior data is associated with the user ID, and then the retrieval data can be analyzed based on NLP natural language processing algorithms to obtain the behavior data benchmark library. Then, based on the behavior data benchmark library and the Euclidean distance algorithm, the user benchmark library can be obtained. Finally, the data popularity library can be calculated based on the user benchmark library and the behavior data benchmark library.

[0016] In an optional implementation, the step of analyzing the retrieved data based on NLP (Natural Language Processing) algorithms to obtain the behavioral data benchmark database includes:

[0017] The retrieved data is used as the training set data, and Chinese word segmentation is performed on the training set data to obtain the word segmentation results;

[0018] The stop word filtering model is used to remove the words that have no value from the word segmentation results, and the target text is obtained.

[0019] The target text is tagged based on the encoder, and a data value matrix is ​​obtained;

[0020] The data value matrix is ​​used as input to a multinomial Naive Bayes model so that the multinomial Naive Bayes model can perform text classification and recognition on the target text and determine the accuracy of the text classification and recognition results. The multinomial Naive Bayes model is used to classify and recognize the target text as either a user-focused instance or user-focused instance data.

[0021] The behavioral data benchmark library is formed based on the text classification and recognition results and the accuracy of the text classification and recognition results.

[0022] This optional implementation uses the retrieved data as training data and performs Chinese word segmentation on the training data to obtain segmentation results. Then, a stop word filtering model is used to remove worthless segments from the segmentation results to obtain the target text. The target text is then tagged using an encoder to obtain a data value matrix. This data value matrix is ​​then used as input to a multinomial Naive Bayes model, enabling the model to perform text classification and recognition on the target text and determine the accuracy of the text classification results. The multinomial Naive Bayes model is used to classify the target text as either a user-focused instance or user-focused instance data. Based on the text classification results and their accuracy, the behavioral data benchmark library is formed.

[0023] In an optional implementation, obtaining the user benchmark library based on the behavioral data benchmark library and the Euclidean distance algorithm includes:

[0024] Based on the Euclidean distance algorithm, the distance between vectors in the behavioral data benchmark database is calculated to obtain the behavioral similarity calculation results.

[0025] Based on the behavioral similarity calculation results, users whose behavioral similarity meets the first preset condition are identified, and these users are associated to form the user benchmark database.

[0026] This optional implementation calculates the distance between vectors in the behavioral data benchmark database based on the Euclidean distance algorithm, thereby obtaining behavioral similarity calculation results. Based on the behavioral similarity calculation results, users whose behavioral similarity meets the first preset condition can be identified, and users whose behavioral similarity meets the first preset condition can be associated to form the user benchmark database.

[0027] In an optional implementation, the step of calculating the data popularity database based on the user benchmark database and the behavioral data benchmark database includes:

[0028] Based on the user benchmark library and the behavioral data benchmark library, the frequency of user operation behavior and the frequency of user search behavior are analyzed to obtain the data popularity library.

[0029] This optional implementation, based on the user benchmark library and the behavioral data benchmark library, can analyze the frequency of user operation behavior and the frequency of user search behavior, and thus access the data popularity library.

[0030] Secondly, this application provides a user behavior prediction method, the method comprising:

[0031] When a target user is detected to have logged into the system, a search behavior prompt is output based on the behavior data benchmark library, the user benchmark library, and the data popularity library. The behavior data benchmark library, the user benchmark library, and the data popularity library are obtained based on the method described in any of the foregoing embodiments.

[0032] The user behavior prediction method of the second aspect of this application can proactively prompt users for data of interest by utilizing the behavior data benchmark library, the user benchmark library, and the data popularity library, thereby reducing user input search behavior, improving user business scenario experience, enhancing the value and efficiency of behavior data utilization, and thus improving user experience.

[0033] Thirdly, this application provides a user behavior analysis device, the device comprising:

[0034] The acquisition module is used to acquire users' historical behavior data;

[0035] The analysis module is used to obtain analysis results based on the user's historical behavior data, and to form a behavior data benchmark library, a user benchmark library, and a data popularity library based on the analysis results. The user benchmark library includes users who have the same data and business interests.

[0036] The apparatus of the third aspect of this application, by executing a user behavior analysis method, can acquire historical user behavior data, and then obtain analysis results based on the historical user behavior data. Based on the analysis results, it forms a behavior data benchmark library, a user benchmark library, and a data popularity library. The user benchmark library includes users with similar data and business interests. Compared with existing technologies, this application can form a behavior data benchmark library, a user benchmark library, and a data popularity library. These libraries can be used to predict users' continued behavior, thereby fully utilizing historical user behavior data and enhancing its value. Existing technologies can only statistically analyze user data, offering limited usable data and failing to analyze the continuity of users' historical behavior, thus limiting their applicability. Furthermore, existing technologies have a relatively simple approach to recording and analyzing user behavior, resulting in insufficient support for user business and inadequate utilization of data value.

[0037] Fourthly, this application provides a user behavior prediction device, the device comprising:

[0038] The prediction module is used to output search behavior prompts based on a behavior data benchmark library, a user benchmark library, and a data popularity library when a target user is detected to have logged into the system. The behavior data benchmark library, the user benchmark library, and the data popularity library are obtained based on the method described in any of the foregoing embodiments.

[0039] The apparatus of the fourth aspect of this application, by executing a user behavior analysis method, can proactively prompt users for data of interest using a behavior data benchmark library, the user benchmark library, and the data popularity library, thereby reducing user input search behavior, improving user experience in business scenarios, enhancing the value and efficiency of behavior data utilization, and ultimately improving user experience.

[0040] Fifthly, this application provides an electronic device, comprising:

[0041] Processor; and

[0042] The memory is configured to store machine-readable instructions that, when executed by the processor, perform the user behavior analysis method and the user behavior prediction method as described in any of the foregoing embodiments.

[0043] The electronic device of the fifth aspect of this application, by executing a user behavior analysis method and a user behavior prediction method, can obtain a behavior data benchmark library, the user benchmark library, and the data popularity library, thereby making full use of historical user behavior data, and can proactively prompt users for data of interest using the behavior data benchmark library, the user benchmark library, and the data popularity library, reducing user input search behavior, improving user business scenario experience, improving the utilization value and efficiency of behavior data, and thus improving user experience.

[0044] Sixthly, this application provides a storage medium storing a computer program, which is executed by a processor using the user behavior analysis method and the user behavior prediction method described in any of the foregoing embodiments.

[0045] The storage medium of the sixth aspect of this application, by executing a user behavior analysis method and a user behavior prediction method, can obtain a behavior data benchmark library, the user benchmark library, and the data popularity library, thereby making full use of historical user behavior data, and can proactively prompt users for data of interest using the behavior data benchmark library, the user benchmark library, and the data popularity library, reducing user input search behavior, improving user business scenario experience, improving the utilization value and efficiency of behavior data, and thus improving user experience. Attached Figure Description

[0046] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart illustrating a user behavior analysis method disclosed in an embodiment of this application;

[0048] Figure 2 This is a flowchart illustrating a user behavior prediction method disclosed in an embodiment of this application;

[0049] Figure 3 This is a schematic diagram of a user behavior analysis and prediction framework disclosed in an embodiment of this application;

[0050] Figure 4 This is a schematic diagram of the structure of a user behavior analysis device disclosed in an embodiment of this application;

[0051] Figure 5 This is a schematic diagram of the structure of a user behavior prediction device disclosed in an embodiment of this application;

[0052] Figure 6 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. Detailed Implementation

[0053] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0054] Example 1

[0055] Please see Figure 1 , Figure 1 This is a flowchart illustrating a user behavior analysis method disclosed in an embodiment of this application, such as... Figure 1 As shown, the method in this application embodiment includes the following steps:

[0056] 101. Obtain user historical behavior data;

[0057] 102. Analyze the results based on historical user behavior data, and form a behavior data benchmark library, a user benchmark library, and a data popularity library based on the analysis results. The user benchmark library includes users who have the same data and business interests.

[0058] This application's embodiments can acquire users' historical behavior data, and then obtain analysis results based on this data. Based on these results, a behavior data benchmark library, a user benchmark library, and a data popularity library are formed. The user benchmark library includes users who share similar data and business interests. Compared to existing technologies, this application can form behavior data benchmark libraries, user benchmark libraries, and data popularity libraries, which can be used to predict users' continued behavior. This allows for full utilization of historical user behavior data and enhances its value. Existing technologies can only statistically analyze user data, offering limited usable data and failing to analyze the continuity of users' historical behavior, thus limiting their applicability. Furthermore, existing technologies have a relatively simple approach to recording and analyzing user behavior, resulting in insufficient support for user businesses and inadequate utilization of data value.

[0059] In this embodiment of the application, the behavioral data benchmark library is used as the basis for predicting the user's subsequent behavior and outputting prompt content. For example, assuming that the behavioral data benchmark library corresponding to the user is: {threat, APT attack}, then "threat" and "APT attack" can be used as prompt content.

[0060] In this embodiment, the user baseline database includes users who share the same data and services they follow, serving as a dependency for determining related users. For example, if user A follows the network threat service, and user B also follows the network threat service, then user B is related to user A, and the user baseline database stores the association between user A and user B. Accordingly, when generating search behavior prompts, the behavior data corresponding to user B can also be used as search behavior prompts. For example, assuming user A follows service S1, and user B follows services S1 and S2, after user A logs into the system, since user A is associated with user B, the search behavior prompts output to user A will include services S1 and S2.

[0061] In this application embodiment, the data popularity database stores the frequency of word segmentation. For example, for the search keyword G, the frequency is 10 times / week, and for the search keyword S, the frequency is 5 times / day.

[0062] In an optional implementation, the steps of obtaining analysis results based on user historical behavior data, and forming a behavior data benchmark library, a user benchmark library, and a data popularity library based on the analysis results, include the following sub-steps:

[0063] Identify the operation data and retrieval data in the user's historical behavior data, use the operation data and retrieval data as the original behavior data, and associate the original behavior data with the user ID;

[0064] Based on NLP (Natural Language Processing) algorithms, retrieved data was analyzed and a behavioral data benchmark library was obtained.

[0065] A user benchmark library is obtained based on a behavioral data benchmark library and the Euclidean distance algorithm;

[0066] A data popularity database is calculated based on a user benchmark database and a behavioral data benchmark database.

[0067] This optional implementation can identify operation data and retrieval data in user historical behavior data, use operation data and retrieval data as raw behavior data, associate raw behavior data with user ID, and then analyze retrieval data based on NLP natural language processing algorithm to obtain behavior data benchmark library, and then obtain user benchmark library based on behavior data benchmark library and Euclidean distance algorithm, and then calculate data popularity library based on user benchmark library and behavior data benchmark library.

[0068] In the above optional implementation, the user ID can be obtained after the user logs into the system.

[0069] In the above optional implementation, a specific way to identify operation data and retrieval data in user historical behavior data is as follows: records such as selecting a business function menu, selecting a certain data, or clicking a button are identified as operation data, and records such as input boxes, drop-down boxes, radio buttons, check boxes, and data list views are identified as retrieval data.

[0070] In an optional implementation, the step of analyzing and retrieving data based on NLP (Natural Language Processing) algorithms to obtain a behavioral data benchmark database includes the following sub-steps:

[0071] Use the retrieved data as the training set data, and perform Chinese word segmentation on the training set data to obtain the word segmentation result. Here, the training set data is a Chinese sentence composed of multiple words. Therefore, it is necessary to split this Chinese sentence into multiple words. For example, for the Chinese sentence "network threat", its corresponding word segmentation result can be "network" and "threat".

[0072] Based on the stop word filtering model, remove the worthless word segments in the word segmentation result and obtain the target text. Here, the stop word filtering model can remove worthless word segments such as "de" (的 in Chinese).

[0073] Based on the encoder, perform labeling processing on the target text and obtain the data value data matrix. Here, the encoder can be a one-hot encoder. The one-hot encoder can encode each word segment in the target text into a one-hot encoding. Furthermore, the one-hot encodings of multiple word segments can form the data value data matrix.

[0074] Use the data value data matrix as the input of the multinomial naive Bayes model to enable the multinomial naive Bayes model to perform text classification recognition on the target text and determine the accuracy of the text classification recognition result. Here, the multinomial naive Bayes model is used to classify the target text into one of the user - concerned instances and the user - concerned instance data. The user - concerned instance is the business that the user is concerned about, and the user - concerned instance data is the data of the business that the user is concerned about. For example, the user - concerned instance can be "network threat", and the user - concerned instance data can be the specific data A and specific data B under the "network threat" business. Further, since multiple users who are concerned about the same business have different specific business data, it is necessary to use the multinomial naive Bayes model to perform text classification recognition on the target text.

[0075] Based on the text classification recognition result and the accuracy of the text classification recognition result, form a behavior data benchmark library. Specifically, the text classification recognition result with high accuracy can be used as the behavior data benchmark library.

[0076] In this optional implementation, by using the retrieved data as the training set data and performing Chinese word segmentation on the training set data, the word segmentation result can be obtained. Furthermore, based on the stop word filtering model, the worthless word segments in the word segmentation result can be removed to obtain the target text. Then, based on the encoder, the target text can be labeled and the data value data matrix can be obtained. Next, the data value data matrix can be used as the input of the multinomial naive Bayes model to enable the multinomial naive Bayes model to perform text classification recognition on the target text and determine the accuracy of the text classification recognition result. Here, the multinomial naive Bayes model is used to classify the target text into one of the user - concerned instances and the user - concerned instance data. Finally, based on the text classification recognition result and the accuracy of the text classification recognition result, the behavior data benchmark library can be formed.

[0077] In an optional implementation, the step of obtaining a user benchmark library based on a behavioral data benchmark library and an Euclidean distance algorithm includes the following sub-steps:

[0078] Based on the Euclidean distance algorithm, the distance between vectors in the behavioral data benchmark database is calculated to obtain the behavioral similarity calculation results.

[0079] Based on the behavioral similarity calculation results, users whose behavioral similarity meets the first preset condition are identified, and these users are associated to form a user benchmark database.

[0080] This optional implementation calculates the distance between vectors in the behavioral data benchmark database based on the Euclidean distance algorithm, which can obtain the behavioral similarity calculation results. Then, based on the behavioral similarity calculation results, it can determine users whose behavioral similarity meets the first preset condition, and associate users whose behavioral similarity meets the first preset condition to form a user benchmark database.

[0081] In the above optional implementation, as an example, a behavioral data benchmark database based on NLP natural language processing algorithms is used to calculate the distance between vectors using the Euclidean Distance algorithm, where it is assumed that the distance between two points in the same dimensional space is: Furthermore, since the Euclidean distance is a non-negative number with a maximum value of positive infinity, the expected value of the similarity calculation is between [-1,1] or [0,1]. Therefore, it is generally used as: sim(p,q)=1 / (1+E(p,q)).

[0082] In an optional implementation, the step of calculating a data popularity database based on a user benchmark database and a behavioral data benchmark database includes the following sub-steps:

[0083] Based on user benchmark databases and behavioral data benchmark databases, the frequency of user operation behaviors and the frequency of user search behaviors are analyzed to obtain a data popularity database.

[0084] This optional implementation is based on a user benchmark library and a behavioral data benchmark library, which can analyze the frequency of user operation behavior and the frequency of user search behavior, and thus obtain a data popularity library.

[0085] In this embodiment, as an example, after users A and B log in to the system, their detailed operations are recorded and submitted to the behavior recognition manager. The behavior recognition manager categorizes and manages the behaviors of users A and B, classifying data related to function menu operations as operation data and data related to data retrieval and query as retrieval data, and stores this operation data and retrieval data as raw behavior data. The retrieval data is segmented using NLP (Natural Language Processing) technology to identify its final retrieval target. The results of the behavior recognition analysis of users A and B are synchronized to the behavior data benchmark library, and users A and B are respectively tagged with corresponding behavior data benchmark library labels. Through the data in the behavior data benchmark library and by performing similarity analysis using its own data, the behavior data of different users A and B can be associated. Specifically, when data with close similarity is found, the associated behavior data benchmark library labels for users A and B are updated, so that the overlapping behaviors of users A and B are unified, that is, the points of interest of users A and B are similar to those of users B, and thus, the expected target results of users A and B are determined. Furthermore, the behavior analysis engine performs frequency analysis on the segmented data and sorts it from high to low frequency. Through multiple iterations of this process, after a user logs in, the search behavior prompter can prioritize prompts for user actions and search data based on user information. This improves the user's experience in business scenarios, enhances the value and efficiency of behavior data utilization, and ultimately improves the user experience.

[0086] Example 2

[0087] Please see Figure 2 , Figure 2 This is a flowchart illustrating a user behavior prediction method disclosed in an embodiment of this application, as shown below. Figure 2 As shown, the method in this application embodiment includes the following steps:

[0088] 201. When a target user is detected to have logged into the system, a search behavior prompt is output based on the behavior data benchmark library, the user benchmark library, and the data popularity library, wherein the behavior data benchmark library, the user benchmark library, and the data popularity library are obtained based on any of the methods described in the foregoing embodiments.

[0089] The user behavior prediction method in this application can proactively prompt users for data they care about by utilizing a behavior data benchmark library, a user benchmark library, and a data popularity library, thereby reducing user input search behavior, improving user experience in business scenarios, enhancing the value and efficiency of behavior data utilization, and ultimately improving user experience.

[0090] In the embodiments of this application, please refer to Figure 3 , Figure 3 This is a schematic diagram of a user behavior analysis and prediction framework disclosed in an embodiment of this application. For example... Figure 3As shown, search behavior suggestions can be output based on the user benchmark library, the data benchmark library (behavioral data benchmark library), and the data popularity library.

[0091] It should be noted that, in addition to generating search behavior prompts, the behavioral data benchmark library, user benchmark library, and data popularity library can also be applied to other scenarios, such as recommending other content.

[0092] Example 3

[0093] Please see Figure 4 , Figure 4 This is a schematic diagram of the structure of a user behavior analysis device disclosed in an embodiment of this application, as shown below. Figure 4 As shown, the apparatus in this embodiment includes the following functional modules:

[0094] Module 301 is used to acquire user historical behavior data;

[0095] Analysis module 302 is used to obtain analysis results based on users' historical behavior data, and to form a behavior data benchmark library, a user benchmark library and a data popularity library based on the analysis results. The user benchmark library includes users who have the same data and business interests.

[0096] The apparatus in this application embodiment, by executing a user behavior analysis method, can acquire historical user behavior data, and then obtain analysis results based on this data. Based on these results, it forms a behavior data benchmark library, a user benchmark library, and a data popularity library. The user benchmark library includes users with similar data and business interests. Compared to existing technologies, this application can form behavior data benchmark libraries, user benchmark libraries, and data popularity libraries, which can be used to predict users' continued behavior. This allows for full utilization of historical user behavior data and enhances its value. Existing technologies can only statistically analyze user data, offering limited usable data and failing to analyze the continuity of users' historical behavior, thus limiting their applicability. Furthermore, existing technologies have a relatively simple approach to recording and analyzing user behavior, resulting in insufficient support for user businesses and inadequate utilization of data value.

[0097] It should be noted that for other detailed descriptions of the apparatus in the embodiments of this application, please refer to the relevant description in Embodiment 1 of this application, which will not be repeated in the embodiments of this application.

[0098] Example 4

[0099] Please see Figure 5 , Figure 5 This is a schematic diagram of the structure of a user behavior prediction device disclosed in an embodiment of this application, as shown below. Figure 5As shown, the apparatus in this embodiment includes the following functional modules:

[0100] The prediction module 401 is used to output search behavior prompts based on the behavior data benchmark library, the user benchmark library and the data popularity library after detecting that a target user has logged into the system. The behavior data benchmark library, the user benchmark library and the data popularity library are obtained based on the method of any of the aforementioned embodiments.

[0101] The apparatus in this application embodiment executes a user behavior analysis method, thereby proactively prompting users for data of interest using a behavior data benchmark library, a user benchmark library, and a data popularity library. This reduces user input search behavior, improves user experience in business scenarios, enhances the value and efficiency of behavior data utilization, and ultimately improves user experience.

[0102] It should be noted that for other detailed descriptions of the apparatus in the embodiments of this application, please refer to the relevant description in Embodiment 1 of this application, which will not be repeated in the embodiments of this application.

[0103] Example 5

[0104] Please see Figure 6 , Figure 6 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. For example... Figure 6 As shown, the electronic device in this application embodiment includes:

[0105] Processor 501; and

[0106] The memory 502 is configured to store machine-readable instructions that, when executed by the processor 501, perform the user behavior analysis method and the user behavior prediction method as described in any of the foregoing embodiments.

[0107] The electronic device in this application embodiment can obtain a behavior data benchmark library, a user benchmark library, and a data popularity library by executing a user behavior analysis method and a user behavior prediction method. This allows for full utilization of historical user behavior data. Furthermore, it can proactively prompt users for data they care about using the behavior data benchmark library, user benchmark library, and data popularity library, thereby reducing user input search behavior, improving user experience in business scenarios, enhancing the value and efficiency of behavior data utilization, and ultimately improving user experience.

[0108] Example 6

[0109] This application provides a storage medium storing a computer program, which is executed by a processor using the user behavior analysis method and the user behavior prediction method described in any of the foregoing embodiments.

[0110] The electronic device in this application embodiment can obtain a behavior data benchmark library, a user benchmark library, and a data popularity library by executing a user behavior analysis method and a user behavior prediction method. This allows for full utilization of historical user behavior data. Furthermore, it can proactively prompt users for data they care about using the behavior data benchmark library, user benchmark library, and data popularity library, thereby reducing user input search behavior, improving user experience in business scenarios, enhancing the value and efficiency of behavior data utilization, and ultimately improving user experience.

[0111] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interface; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0112] Furthermore, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0113] Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0114] It should be noted that if a function is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0115] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0116] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A user behavior analysis method, characterized in that, The method includes: Obtain user historical behavior data; The analysis results are obtained based on the user's historical behavior data, and a behavior data benchmark library, a user benchmark library, and a data popularity library are formed based on the analysis results. The user benchmark library includes users who have the same data and business interests. Furthermore, the process of obtaining analysis results based on the user's historical behavior data, and forming a behavior data benchmark library, a user benchmark library, and a data popularity library based on the analysis results, includes: Identify the operation data and retrieval data in the user's historical behavior data, and use the operation data and retrieval data as raw behavior data, and associate the raw behavior data with the user ID; The retrieved data is analyzed based on NLP (Natural Language Processing) algorithms to obtain the behavioral data benchmark library; The user benchmark library is obtained based on the aforementioned behavioral data benchmark library and the Euclidean distance algorithm; The data heat library is calculated based on the user benchmark library and the behavioral data benchmark library; And, the step of analyzing the retrieved data based on NLP (Natural Language Processing) algorithms to obtain the behavioral data benchmark library includes: The retrieved data is used as the training set data, and Chinese word segmentation is performed on the training set data to obtain the word segmentation results; The stop word filtering model is used to remove the words that have no value from the word segmentation results, and the target text is obtained. The target text is tagged based on the encoder, and a data value matrix is ​​obtained; The data value matrix is ​​used as input to a multinomial Naive Bayes model so that the multinomial Naive Bayes model can perform text classification and recognition on the target text and determine the accuracy of the text classification and recognition results. The multinomial Naive Bayes model is used to classify and recognize the target text as either a user-focused instance or user-focused instance data. The behavioral data benchmark library is formed based on the text classification and recognition results and the accuracy of the text classification and recognition results.

2. The method as described in claim 1, characterized in that, The user benchmark library, obtained based on the behavioral data benchmark library and the Euclidean distance algorithm, includes: Based on the Euclidean distance algorithm, the distance between vectors in the behavioral data benchmark database is calculated to obtain the behavioral similarity calculation results. Based on the behavioral similarity calculation results, users whose behavioral similarity meets the first preset condition are identified, and these users are associated to form the user benchmark database.

3. The method as described in claim 1, characterized in that, The process of calculating the data popularity database based on the user benchmark database and the behavioral data benchmark database includes: Based on the user benchmark library and the behavioral data benchmark library, the frequency of user operation behavior and the frequency of user search behavior are analyzed to obtain the data popularity library.

4. A user behavior prediction method, characterized in that, The method includes: When a target user is detected to have logged into the system, a search behavior prompt is output based on the behavior data benchmark library, the user benchmark library, and the data popularity library, wherein the behavior data benchmark library, the user benchmark library, and the data popularity library are obtained based on the method described in any one of claims 1-3.

5. A user behavior analysis device, characterized in that, The device includes: The acquisition module is used to acquire users' historical behavior data; The analysis module is used to obtain analysis results based on the user's historical behavior data, and to form a behavior data benchmark library, a user benchmark library, and a data popularity library based on the analysis results. The user benchmark library includes users who have the same data and business interests. Furthermore, the process of obtaining analysis results based on the user's historical behavior data, and forming a behavior data benchmark library, a user benchmark library, and a data popularity library based on the analysis results, includes: Identify the operation data and retrieval data in the user's historical behavior data, and use the operation data and retrieval data as raw behavior data, and associate the raw behavior data with the user ID; The retrieved data is analyzed based on NLP (Natural Language Processing) algorithms to obtain the behavioral data benchmark library; The user benchmark library is obtained based on the aforementioned behavioral data benchmark library and the Euclidean distance algorithm; The data heat library is calculated based on the user benchmark library and the behavioral data benchmark library; And, the step of analyzing the retrieved data based on NLP (Natural Language Processing) algorithms to obtain the behavioral data benchmark library includes: The retrieved data is used as the training set data, and Chinese word segmentation is performed on the training set data to obtain the word segmentation results; The stop word filtering model is used to remove the words that have no value from the word segmentation results, and the target text is obtained. The target text is tagged based on the encoder, and a data value matrix is ​​obtained; The data value matrix is ​​used as input to a multinomial Naive Bayes model so that the multinomial Naive Bayes model can perform text classification and recognition on the target text and determine the accuracy of the text classification and recognition results. The multinomial Naive Bayes model is used to classify and recognize the target text as either a user-focused instance or user-focused instance data. The behavioral data benchmark library is formed based on the text classification and recognition results and the accuracy of the text classification and recognition results.

6. A user behavior prediction device, characterized in that, The device includes: The prediction module is used to output search behavior prompts based on a behavior data benchmark library, a user benchmark library, and a data popularity library when a target user is detected to have logged into the system. The behavior data benchmark library, the user benchmark library, and the data popularity library are obtained based on the method described in any one of claims 1-3.

7. An electronic device, characterized in that, include: processor; as well as The memory is configured to store machine-readable instructions that, when executed by the processor, perform the user behavior analysis method as described in any one of claims 1-3 and the user behavior prediction method as described in claim 4.

8. A storage medium, characterized in that, The storage medium stores a computer program, which is executed by a processor using the user behavior analysis method as described in any one of claims 1-3 and the user behavior prediction method as described in claim 4.

Images