User feature generation method and system, electronic equipment and storage medium

Abstract

The invention discloses a user feature generation method and system, electronic equipment and a storage medium. The method comprises the steps that behavior data generated in the process that a user uses a target platform is collected; identifying a behavior pattern of the user based on the behavior data; and generating dynamic user characteristics based on the behavior pattern of the user. According to the method, rich user preference information contained in the user behavior data is fully utilized, the behavior characteristics and trends of the users can be accurately analyzed, the dynamic habits and search preferences of the users can be accurately reflected, accurate search results can be provided for the users, the search cost of the users is greatly reduced, and the search efficiency is improved. In addition, customized project recommendation can be provided, the matching efficiency is improved, the service quality and the operation efficiency of the target platform are improved, and the participation degree and the satisfaction degree of the user are improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a user feature generation method and system, electronic equipment, and a computer-readable storage medium. Background Art

[0002] With the continuous development of Internet technology, many platforms, in order to improve service quality and user experience, analyze users' characteristics and recommend personalized and accurate services to users based on user characteristics. For example, in the field of bidding, the recommendation system, as an efficient information matching tool, is of great significance for improving the matching efficiency of tenderers and bidders and optimizing resource allocation. Traditional recommendation systems mostly rely on static user attributes and historical transaction data to generate static user features, but in a rapidly changing market environment, this static feature generation method gradually shows its limitations, which seriously affects the matching efficiency. For example, in the bidding process, user needs and preferences may change rapidly with time and project characteristics, and traditional methods are difficult to capture this dynamic, which makes it difficult for users to accurately match suitable projects. For example, Chinese patent application CN113220999A discloses a method for generating user features, which obtains historical text data corresponding to the target user, parses the data, determines the word set and the number of word segments under the theme, and thus generates user features, but this method is only based on static historical text data for analysis, and can only generate static user features, but cannot generate dynamic user features. Summary of the invention

[0003] The present invention provides a user feature generation method and system, an electronic device, and a computer-readable storage medium, which can generate dynamic user features, accurately reflect the user's dynamic habits and search preferences, and are conducive to providing users with accurate search results, improving matching efficiency, and enhancing the service quality and operational efficiency of the target platform.

[0004] According to one aspect of the present invention, a method for generating user characteristics is provided, comprising the following contents:

[0005] Collect behavioral data generated by users when using the target platform;

[0006] identifying a user's behavior pattern based on the behavior data;

[0007] Generating dynamic user features based on the user's behavior pattern specifically includes: identifying potential features related to user preferences and needs from the user's behavior pattern; combining the identified potential features to construct characteristic features of the deep-level behavior pattern; extracting time series features from the behavior data sequence; evaluating the importance of potential features, characteristic features and time series features and sorting the features from high to low in terms of importance, and selecting at least one feature with a high ranking as the dynamic user feature.

[0008] Furthermore, behavioral data includes browsing behavior data, search query data, personalization setting data, and device data.

[0009] Furthermore, the process of identifying the user's behavior pattern based on the behavior data includes the following:

[0010] The user's behavior data is constructed into a behavior data sequence in chronological order, and the sequence mining algorithm is used to mine the continuous actions of the user including setting filtering conditions for specific keywords, search queries, browsing, and collecting from the behavior data sequence.

[0011] Furthermore, the following contents are also included:

[0012] The context information corresponding to the dynamic user features is extracted, and the weight of each dynamic user feature is adjusted according to the context information.

[0013] Furthermore, the following contents are also included:

[0014] Build a recommendation model based on dynamic user features, evaluate the effectiveness of the recommendation model, and optimize the dynamic user features based on the evaluation results.

[0015] Furthermore, after collecting the behavioral data generated by the user in the process of using the target platform, the following contents are also included:

[0016] The collected behavioral data are preprocessed to obtain unified and standardized behavioral data.

[0017] In addition, the present invention also provides a user feature generation system, comprising:

[0018] The behavior data collection module is used to collect the behavior data generated by users when using the target platform;

[0019] A behavior pattern recognition module, used to recognize the user's behavior pattern based on the behavior data;

[0020] The dynamic user feature generation module is used to generate dynamic user features based on the user's behavior pattern, specifically for: identifying potential features related to user preferences and needs from the user's behavior pattern; combining the identified potential features to construct characteristic features of deep-level behavior patterns; extracting time series features from the behavior data sequence; evaluating the importance of potential features, characteristic features and time series features and sorting the features from high to low in importance, and selecting at least one feature with a high ranking as a dynamic user feature.

[0021] In addition, the present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the above method by calling the computer program stored in the memory.

[0022] In addition, the present invention also provides a computer-readable storage medium for storing a computer program for generating dynamic user features based on user behavior, wherein the computer program executes the steps of the method described above when running on a computer.

[0023] The present invention has the following beneficial effects:

[0024] The user feature generation method of the present invention collects the behavior data generated by the user in the process of using the target platform, then identifies the user's behavior pattern based on the behavior data, and finally generates dynamic user features based on the user's behavior pattern. The present invention makes full use of the rich user preference information contained in the user behavior data, which is conducive to accurately analyzing the user's behavior characteristics and trends, can accurately reflect the user's dynamic habits and search preferences, and is conducive to providing users with accurate search results, greatly reducing the user's search cost, and can also provide customized project recommendations, improve matching efficiency, improve the service quality and operation efficiency of the target platform, and improve user participation and satisfaction.

[0025] In addition, the user feature generation system of the present invention also has the above advantages.

[0026] In addition to the above-described purposes, features and advantages, the present invention has other purposes, features and advantages. The present invention will be further described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The drawings constituting a part of this application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0028] Figure 1 It is a flowchart of a method for generating user characteristics according to a preferred embodiment of the present application;

[0029] Figure 2 yes Figure 1 Schematic diagram of the sub-process of step S3;

[0030] Figure 3 is another flowchart of the user feature generation method according to the preferred embodiment of the present application;

[0031] Figure 4 is another flowchart of the user feature generation method according to the preferred embodiment of the present application;

[0032] Figure 5 is another flowchart of the user feature generation method according to the preferred embodiment of the present application;

[0033] Figure 6 is another flowchart of the user feature generation method according to the preferred embodiment of the present application;

[0034] Figure 7 It is a schematic diagram of the module structure of a user feature generation system according to another embodiment of the present application. DETAILED DESCRIPTION

[0035] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0036] Reference Figure 1 As shown, the preferred embodiment of the present application provides a method for generating user characteristics, comprising the following contents:

[0037] Step S1: Collecting behavioral data generated by users when using the target platform;

[0038] It can be understood that in step S1, the target platform refers to a platform that can collect user behavior data, such as major websites, apps, mini-programs, etc., to collect the behavior data generated by users when using the target platform. The behavior data includes browsing behavior data, search query data, personalized settings data and device data.

[0039] Specifically, taking the collection of user behavior data on the bidding platform as an example, the browsing behavior data includes data such as browsed items, dwell time, visited pages, and the user's interest in specific items. The search query data includes data such as the keywords entered by the user in the search box and the click-through rate of the search results. The search query data is conducive to revealing the user's query intentions and query preferences. The personalized setting data includes data such as the industry fields that the user is concerned about and the project types that he prefers, which helps to provide customized recommendations for users. The device data includes data such as the device model and geographic location information of the device used by the user.

[0040] Step S2: Identifying the user's behavior pattern based on the behavior data;

[0041] It can be understood that in step S2, the process of identifying the user's behavior pattern based on the behavior data includes the following:

[0042] The user's behavior data is constructed into a behavior data sequence in chronological order, and the sequence mining algorithm is used to mine the continuous actions of the user including setting filtering conditions for specific keywords, search queries, browsing, and collecting from the behavior data sequence.

[0043] Specifically, the user behavior data collected in step S1 is constructed into a behavior data sequence in chronological order, and then a sequence mining algorithm such as a hidden Markov model is used to mine the continuous actions of the user setting filtering conditions for specific keywords, search queries, browsing, and collecting from the behavior data sequence, so that the behavior patterns and trends in the user behavior can be discovered, which is conducive to accurately analyzing the user's behavioral characteristics and trends.

[0044] Step S3: Generate dynamic user features based on the user's behavior pattern.

[0045] It is understandable that Figure 2 As shown, in step S3, the process of generating dynamic user features based on the user's behavior pattern includes the following contents:

[0046] Step S3.1: Identify potential features related to user preferences and needs from user behavior patterns;

[0047] Step S3.2: Combining the identified latent features to construct characteristic features of the deep-level behavior pattern;

[0048] Step S3.3: extracting temporal features from the behavioral data sequence;

[0049] Step S3.4: Evaluate the importance of potential features, characteristic features and time series features and sort the features from high to low in importance, and select at least one feature with a high ranking as a dynamic user feature.

[0050] It can be understood that, based on business understanding and data analysis, potential features related to user preferences and needs are first identified from the user's continuous actions of setting filtering conditions for specific keywords, search queries, browsing, and collections, and then the identified potential features are combined to construct characteristic features of deep-level behavior patterns, and then the time series features are extracted from the behavior data sequence based on time series processing technology. Finally, the importance of potential features, characteristic features, and time series features is evaluated and the features are sorted from large to small according to importance, and at least one feature with a high ranking is selected as a dynamic user feature. Among them, the specific number of dynamic user features can be set according to actual needs and is not specifically limited here. In addition, the generated dynamic user features can also be stored and managed, and a feature library can be established to facilitate model training and subsequent recommendation system use. In addition, the potential features and characteristic features can also be mathematically transformed to enhance the performance of the model, such as standardizing numerical features, encoding categorical features, and using the dimensionality reduction technology of principal component analysis (PCA) to reduce the dimensionality of data.

[0051] Specifically, taking the bidding platform as an example, firstly, based on business understanding and data analysis, the potential features related to user preferences and needs are identified from the continuous actions of users setting filtering conditions for specific keywords, searching, browsing, and collecting, including the business scope, region, project type, project amount, purchaser type, etc. that users are concerned about. Secondly, the identified potential features are combined to construct the characteristic features of the deep-level behavior pattern. For example, the characteristic feature of "market share" is constructed by combining the user's winning bid record and all projects in the field. Then, dynamic time series features are extracted from the behavioral data sequence using time window, sliding average and other technologies. Finally, the importance of potential features, characteristic features and time series features is evaluated. For example, the feature selection method of statistical test is used to evaluate the importance of each feature to the prediction target, so that the most influential (i.e., most important) features can be identified, and the stability of features over time can be evaluated to ensure that the features remain consistent in different time periods to improve the generalization ability of the model, and at least one feature with a high ranking is selected as a dynamic user feature, for example, the most important feature is selected as a dynamic user feature.

[0052] It can be understood that the bidding user feature generation method based on user behavior in this embodiment collects the behavior data generated by the user in the process of using the target platform, then identifies the user's behavior pattern based on the behavior data, and finally generates dynamic user features based on the user's behavior pattern. The present invention makes full use of the rich user preference information contained in the user behavior data, which is conducive to accurately analyzing the user's behavior characteristics and trends, can accurately reflect the user's dynamic habits and search preferences, and is conducive to providing users with accurate search results, greatly reducing the user's search costs, and can also provide customized project recommendations, improve matching efficiency, improve the service quality and operational efficiency of the target platform, and improve user participation and satisfaction.

[0053] like Figure 3 As shown, the user feature generation method also includes the following contents:

[0054] Step S4: extracting context information corresponding to the dynamic user features, and adjusting the weight of each dynamic user feature according to the context information.

[0055] It can be understood that in the complex environment of the bidding platform, the generation of dynamic user features not only depends on the user's behavioral data, but also needs to consider rich contextual information. Combining user behavior with environmental factors can provide more comprehensive and accurate personalized services. Therefore, the present invention also extracts contextual information corresponding to dynamic user features, such as time, location, seasonal trends of projects, special requirements of the industry, etc., and dynamically adjusts the weight and influence of each dynamic user feature according to the contextual information to ensure the timeliness and adaptability of dynamic user features. For example, the region-related features are adjusted in real time according to the user's recent positioning information; for example, some business projects have seasonal trends, and the weights of users and business-related features will change with the seasons.

[0056] like Figure 4 As shown, the user feature generation method also includes the following contents:

[0057] Step S5: Update the dynamic user characteristics in real time.

[0058] It is understandable that in a rapidly changing bidding environment, user needs and behavior patterns may evolve rapidly. To ensure the accuracy and responsiveness of the recommendation system, dynamic user features need to be updated in real time. Therefore, the present invention adopts real-time data stream monitoring technology to continuously track user activities on the bidding platform, and adopts an event-driven architecture. Whenever a user performs a key operation, such as a search operation, the relevant features can be immediately captured and updated to the feature library to ensure real-time updating of the feature library. In addition, time window technology is also used to aggregate and analyze user behavior data in real time to reflect the user's activity pattern in a specific time period.

[0059] like Figure 5 As shown, the user feature generation method also includes the following contents:

[0060] Step S6: construct a recommendation model based on the dynamic user features, evaluate the effect of the recommendation model, and optimize the dynamic user features according to the evaluation results.

[0061] It can be understood that after generating dynamic user features, building a recommendation model based on the dynamic user features can realize personalized recommendation services. Specifically, first clarify the goal of the recommendation model, such as improving user participation, improving user satisfaction, etc., to ensure that the model is built in line with the business goals, and then integrate the generated dynamic user features into the recommendation model as the input of the model, and train the recommendation model, which is conducive to improving the accuracy and personalization of the recommendation model. Among them, the recommendation model belongs to the existing technology, and the specific construction process is not repeated here. Then, according to the business goals and user needs, set key evaluation indicators, such as click-through rate, conversion rate, prediction accuracy, etc., and adopt the A / B testing method, that is, real-time comparison of the effects of different versions of dynamic user feature generation strategies, collect user feedback on the final recommendation results, and evaluate the performance of the generation strategy in actual applications, so as to iteratively optimize the dynamic user feature generation strategy according to the evaluation results, which is conducive to ensuring the efficient operation of the recommendation system, continuously improving user experience, and maximizing business goals.

[0062] like Figure 6 As shown, after collecting the behavior data generated by the user in the process of using the target platform, the user feature generation method also includes the following contents:

[0063] Step S1a: pre-process the collected behavior data to obtain unified and standardized behavior data.

[0064] It can be understood that in order to ensure the quality and availability of user behavior data, the collected behavior data needs to be preprocessed, such as data cleaning, data formatting and data normalization. Among them, data cleaning includes removing duplicate records, invalid data and outliers, data formatting includes unifying the data format to ensure that all behavior records follow the same data structure and type, and standardizing text data, such as case conversion, removal of special characters, etc. Data normalization includes normalizing digital data to eliminate the impact of different dimensions and magnitudes, so as to obtain unified and standardized behavior data.

[0065] like Figure 7 As shown, another embodiment of the present invention further provides a user feature generation system, preferably using the user feature generation method as described above, including:

[0066] The behavior data collection module is used to collect the behavior data generated by users when using the target platform;

[0067] A behavior pattern recognition module, used to recognize the user's behavior pattern based on the behavior data;

[0068] The dynamic user feature generation module is used to generate dynamic user features based on the user's behavior pattern.

[0069] It can be understood that the user feature generation system of this embodiment collects the behavior data generated by the user in the process of using the target platform, identifies the user's behavior pattern based on the behavior data, and finally generates dynamic user features based on the user's behavior pattern. The present invention makes full use of the rich user preference information contained in the user behavior data, which is conducive to accurately analyzing the user's behavior characteristics and trends, can accurately reflect the user's dynamic habits and search preferences, and is conducive to providing users with accurate search results, greatly reducing the user's search costs, and can also provide customized project recommendations, improve matching efficiency, improve the service quality and operational efficiency of the target platform, and improve user participation and satisfaction.

[0070] In addition, the user feature generation system further includes:

[0071] The context information fusion module is used to extract the context information corresponding to the dynamic user features and adjust the weight of each dynamic user feature according to the context information.

[0072] In addition, the user feature generation system further includes:

[0073] The feature real-time update module is used to update dynamic user features in real time.

[0074] In addition, the user feature generation system further includes:

[0075] The recommendation model construction and evaluation optimization module is used to build a recommendation model based on dynamic user features, evaluate the effectiveness of the recommendation model, and optimize the dynamic user features based on the evaluation results.

[0076] In addition, the user feature generation system further includes:

[0077] The data preprocessing module is used to preprocess the collected behavior data to obtain unified and standardized behavior data.

[0078] It can be understood that the various modules of the system embodiment correspond to the various steps of the above method embodiment, so the specific working process of each module will not be repeated here, and the corresponding reference is made to the various steps of the above method embodiment.

[0079] In addition, another embodiment of the present invention further provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the above method by calling the computer program stored in the memory.

[0080] In addition, another embodiment of the present invention further provides a computer-readable storage medium for storing a computer program for generating dynamic user features based on user behavior, wherein the computer program executes the steps of the method described above when running on a computer.

[0081] Common computer-readable storage media include: floppy disks, flexible disks, hard disks, magnetic tapes, any other magnetic media, CD-ROMs, any other optical media, punch cards, paper tapes, any other physical media with patterns of holes, random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), flash-erasable programmable read-only memory (FLASH-EPROM), any other memory chip or cartridge, or any other medium that can be read by a computer. Instructions can further be transmitted or received by a transmission medium. The term transmission medium may include any tangible or intangible medium that can be used to store, encode or carry instructions for execution by a machine, and includes digital or analog communication signals or intangible media that facilitate communication of the above instructions. Transmission media include coaxial cables, copper wire, and optical fiber, including the wires of a bus used to transmit a computer data signal.

[0082] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present application may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0083] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0084] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0085] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0086] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0087] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

[0088] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for generating user characteristics, characterized in that: Includes the following: Collect behavioral data generated by users when using the target platform; identifying a user's behavior pattern based on the behavior data; Generating dynamic user features based on the user's behavior pattern specifically includes: identifying potential features related to user preferences and needs from the user's behavior pattern; Based on the identified latent features, characteristic features are combined to construct characteristic features of deep-level behavioral patterns; time series features are extracted from the behavioral data sequence; the importance of latent features, characteristic features and time series features are evaluated and the features are sorted from high to low in order of importance, and at least one feature with a high ranking is selected as a dynamic user feature.

2. The user feature generation method according to claim 1, characterized in that: Behavioral data includes browsing behavior data, search query data, personalization data, and device data.

3. The user feature generation method according to claim 1, characterized in that: The process of identifying the user's behavior pattern based on the behavior data includes the following: The user's behavior data is constructed into a behavior data sequence in chronological order, and the sequence mining algorithm is used to mine the continuous actions of the user including setting filtering conditions for specific keywords, search queries, browsing, and collecting from the behavior data sequence.

4. The user feature generation method according to claim 1, characterized in that: Also included are the following: The context information corresponding to the dynamic user features is extracted, and the weight of each dynamic user feature is adjusted according to the context information.

5. The user feature generation method according to claim 1, characterized in that: Also included are the following: Build a recommendation model based on dynamic user features, evaluate the effectiveness of the recommendation model, and optimize the dynamic user features based on the evaluation results.

6. The user feature generation method according to claim 1, characterized in that: The collection of behavioral data generated by users when using the target platform also includes the following: The collected behavioral data are preprocessed to obtain unified and standardized behavioral data.

7. A user feature generation system, characterized in that: include: The behavior data collection module is used to collect the behavior data generated by users when using the target platform; A behavior pattern recognition module, used to recognize the user's behavior pattern based on the behavior data; A dynamic user feature generation module, used to generate dynamic user features based on the user's behavior pattern, specifically used to: identify potential features related to user preferences and needs from the user's behavior pattern; Based on the identified latent features, characteristic features are combined to construct characteristic features of deep-level behavioral patterns; time series features are extracted from the behavioral data sequence; the importance of latent features, characteristic features and time series features are evaluated and the features are sorted from high to low in order of importance, and at least one feature with a high ranking is selected as a dynamic user feature.

8. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the method according to any one of claims 1 to 6 by calling the computer program stored in the memory.

9. A computer-readable storage medium for storing a computer program for generating dynamic user features based on user behavior, characterized in that: When the computer program is run on a computer, the steps of the method according to any one of claims 1 to 6 are executed.

Images