User portrait analysis method and device, electronic equipment and storage medium

Abstract

The invention discloses a user portrait analysis method and device, and the method comprises the steps: obtaining an incomplete effective data sample, and enabling the incomplete effective data sample to be requested from a background; obtaining supplementary data samples, wherein the supplementary data samples comprise a static negative sampling data sample, a dynamic negative sampling data sample and a front-end click data sample; calculating distance values between the static negative sampling data sample, the dynamic negative sampling data sample and the front-end click data sample and the incomplete valid data sample; sorting the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples according to an incremental relationship of the distance values; and selecting the first K data samples with the minimum distance value, and counting the occurrence frequency of each sample in the first K data samples. According to the user portrait analysis method and device provided by the invention, the completeness of the user data is greatly improved by supplementing the supplementary data sample to the incomplete effective data sample.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of data processing, and in particular to a method and device for analyzing user portraits. Background Art

[0002] At present, all major financial institutions have big data analysis systems for customer portraits in order to better serve customers. The data analysis obtained is limited to the analysis of valid data, which will ignore the analysis of a large amount of potential customer data, thereby losing customer stickiness.

[0003] The publication number is CN118332197A and the name is "Big Data Push Method and System Based on Digital Finance", which discloses a big data push method based on digital finance, including: obtaining the financial behavior log of the target user; passing the financial behavior log of the target user into the user portrait construction algorithm, and obtaining a first portrait label marking result for each portrait label, the first portrait label marking result characterizing whether the financial behavior log of the target user matches the portrait label; building a user portrait based on the portrait labels matched in the first portrait label marking result, and pushing data based on the user portrait; wherein the user portrait construction algorithm is optimized by the following steps: passing the behavior data template in the initial behavior data template library into multiple user portrait construction algorithms, and obtaining a representative support portrait label corresponding to each user portrait construction algorithm; annotating the first key portrait label of the behavior data template and the representative support corresponding to each user portrait construction algorithm; The consistent behavior data template is matched with the portrait label, and a consistent behavior data template is determined from the initial behavior data template library, and the determined consistent behavior data templates form a consistent behavior data template library; the consistent behavior data template is passed into the user portrait construction algorithm to obtain the first true evaluation support corresponding to each portrait label; the first portrait label annotation information of each portrait label and the first true evaluation support are used to obtain the first error corresponding to each user portrait construction algorithm; for each user portrait construction algorithm, the user portrait construction algorithm is optimized through the first error corresponding to the user portrait construction algorithm; the initial behavior data template library is replaced based on the consistent behavior data template library, and then the behavior data template in the initial behavior data template library is returned to the operation of passing the behavior data template in the initial behavior data template library into multiple user portrait construction algorithms, until the first error reaches the convergence standard, and the optimized user portrait construction algorithm is determined from the multiple user portrait construction algorithms. However, the above method is too complicated and inefficient when used.

[0004] Therefore, it is necessary to provide a method and device for analyzing user portraits to solve the above problems. Summary of the invention

[0005] The purpose of the embodiments of the present application is to propose a method and device for analyzing user portraits, which greatly improves the integrity of user data by supplementing incomplete valid data samples with static negative sampling data samples, dynamic negative sampling data samples, and front-end click data samples.

[0006] According to one aspect of the present invention, a method for analyzing a user portrait is provided, the method comprising: Obtaining an incomplete valid data sample, wherein the incomplete valid data sample is requested from a background; Acquire supplementary data samples, wherein the supplementary data samples include static negative sampling data samples, dynamic negative sampling data samples, and front-end click data samples; Calculating the distance value between the static negative sampling data sample, the dynamic negative sampling data sample, the front-end click data sample and the incomplete valid data sample; Sorting the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples according to the increasing relationship of the distance values; The first K data samples with the smallest distance values ​​are selected, and the number of occurrences of each sample in the first K data samples is counted.

[0007] Preferably, if the distance value is greater than a preset value, the supplementary data sample is discarded, and the distance value is specifically calculated by the following formula: , in, and are the coordinates of the first and second sample points, is the distance value between the first sample point and the second sample point, the first sample point is extracted from the valid data sample, and the second sample point is extracted from the static negative sampling data sample, the dynamic negative sampling data sample, and the front-end click data sample.

[0008] Preferably, the supplementary data sample is data within an activity cycle of the user, and according to the time axis, the time from when the user enters the system to when the user exits the system is regarded as an activity cycle.

[0009] Preferably, the time when a user enters the system and changes the token is regarded as an activity cycle.

[0010] Preferably, a usage cycle of a user on the system is regarded as an activity cycle.

[0011] Preferably, the static negative sampling data samples are selected based on a preset distribution rule; and the dynamic negative sampling data samples are selected based on the performance of the current model and the characteristics of the training data.

[0012] Preferably, the more times each sample appears, the higher the confidence of the corresponding supplementary data sample.

[0013] According to another aspect of the present invention, a user portrait analysis device is provided, the device comprising: A valid data sample acquisition unit, which is used to acquire an incomplete valid data sample, wherein the incomplete valid data sample is requested from a background; A supplementary data sample acquisition unit, which is used to acquire supplementary data samples, wherein the supplementary data samples include static negative sampling data samples, dynamic negative sampling data samples and front-end click data samples; A distance value calculation unit, which is used to calculate the distance value between the static negative sampling data sample, the dynamic negative sampling data sample, the front-end click data sample and the incomplete valid data sample; A data sample sorting unit, configured to sort the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples according to an increasing relationship of the distance values; A data sample statistical unit is used to select the first K data samples with the smallest distance value and count the number of occurrences of each sample in the first K data samples.

[0014] The present application discloses a method and device for analyzing user portraits, which obtain incomplete valid data samples, wherein the incomplete valid data samples are requested from a background; obtain supplementary data samples, wherein the supplementary data samples include static negative sampling data samples, dynamic negative sampling data samples, and front-end click data samples; calculate the distance values ​​between the static negative sampling data samples, the dynamic negative sampling data samples, and the front-end click data samples and the incomplete valid data samples; sort the static negative sampling data samples, the dynamic negative sampling data samples, and the front-end click data samples according to the increasing relationship of the distance values; select the first K data samples with the smallest distance values, and count the number of occurrences of each sample in the first K data samples, so as to greatly improve the integrity of user data by supplementing the incomplete valid data samples with the static negative sampling data samples, the dynamic negative sampling data samples, and the front-end click data samples. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are some embodiments of the present invention, not all embodiments. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor.

[0016] Figure 1 is a flow chart of a method for analyzing a user portrait according to an embodiment of the present invention; Figure 2 is a module schematic diagram of a user portrait analysis device according to an embodiment of the present invention; Figure 3 FIG. 4 is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. According to the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0019] In view of the problems existing in the prior art, an embodiment of the present invention provides a method and device for analyzing user portraits, which greatly improves the integrity of user data by supplementing incomplete valid data samples with static negative sampling data samples, dynamic negative sampling data samples and front-end click data samples.

[0020] Figure 1 FIG. 1 is a flow chart of a method for analyzing a user portrait according to an embodiment of the present invention. Figure 1 According to one aspect of the present invention, a method for analyzing a user portrait is provided, the method comprising: Step S101: obtaining an incomplete valid data sample, wherein the incomplete valid data sample is requested from a background; Step S102: Acquire supplementary data samples, where the supplementary data samples include static negative sampling data samples, dynamic negative sampling data samples, and front-end click data samples; Step S103: calculating the distance value between the static negative sampling data sample, the dynamic negative sampling data sample, the front-end click data sample and the incomplete valid data sample; Step S104: sorting the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples according to the increasing relationship of the distance values; Step S105: Select the first K data samples with the smallest distance value, and count the number of occurrences of each sample in the first K data samples.

[0021] Specifically, based on the data of the valid data samples, the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples are supplemented into the valid data to form effective user usage habits, thereby forming a complete link of operation habits.

[0022] Depending on the business objectives, some data samples are useful and some are useless. For example, if valid data samples are defined as user tracks of users who want to purchase online products, then data samples of users browsing announcements or enjoying rights and interests services are useless and need to be deleted.

[0023] In a specific implementation, if the distance value is greater than a preset value, the supplementary data sample is discarded, and the distance value is specifically calculated by the following formula: , in, and are the coordinates of the first and second sample points, is the distance value between the first sample point and the second sample point, the first sample point is extracted from the valid data sample, and the second sample point is extracted from the static negative sampling data sample, the dynamic negative sampling data sample, and the front-end click data sample.

[0024] In a specific implementation, the supplementary data sample is data within an activity cycle of the user, and according to the time axis, the time when the user enters the system and when the user exits the system is regarded as an activity cycle.

[0025] In the specific implementation, when the user enters the system and changes the token, it is regarded as an activity cycle.

[0026] In a specific implementation, the usage cycle of the user on the system is regarded as an activity cycle.

[0027] In a specific implementation, the static negative sampling data samples are selected based on a preset distribution rule; and the dynamic negative sampling data samples are selected based on the performance of the current model and the characteristics of the training data.

[0028] In a specific implementation, the more times each sample appears, the higher the confidence of the corresponding supplementary data sample.

[0029] Figure 2 is a schematic diagram of a module of a user portrait analysis device according to an embodiment of the present invention. Figure 2According to another aspect of the present invention, a user portrait analysis device is provided, the device comprising: A valid data sample acquisition form, 21, is used to acquire an incomplete valid data sample, the incomplete valid data sample being requested from the background; A supplementary data sample acquisition unit 22, which is used to acquire supplementary data samples, wherein the supplementary data samples include static negative sampling data samples, dynamic negative sampling data samples and front-end click data samples; A distance value calculation unit 23, which is used to calculate the distance value between the static negative sampling data sample, the dynamic negative sampling data sample, the front-end click data sample and the incomplete valid data sample; A data sample sorting unit 24, which is used to sort the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples according to the increasing relationship of the distance values; The data sample statistical unit 25 is used to select the first K data samples with the smallest distance value and count the number of occurrences of each sample in the first K data samples.

[0030] Figure 3 is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application. Figure 3 As shown, an electronic device 300 provided in this embodiment includes: a processor 301 and a memory 302; wherein: The memory 302 is used to store computer programs, and the memory may also be a flash memory.

[0031] The processor 301 is used to execute the execution instructions stored in the memory to implement each step in the above method. For details, please refer to the relevant description in the above method embodiment.

[0032] Optionally, the memory 302 may be independent or integrated with the processor 401 .

[0033] When the memory 302 is a device independent of the processor 301, the electronic device 300 may further include: The bus 303 is used to connect the memory 302 and the processor 301 .

[0034] This embodiment further provides a readable storage medium, in which a computer program is stored. When at least one processor of an electronic device executes the computer program, the electronic device executes the methods provided in the above-mentioned various implementation modes.

[0035] This embodiment also provides a program product, which includes a computer program stored in a readable storage medium. At least one processor of the electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device implements the methods provided in the above various embodiments.

[0036] In summary, the embodiment of the present application discloses a method and device for analyzing user portraits, which obtain incomplete valid data samples, and the incomplete valid data samples are requested from the background; supplementary data samples are obtained, and the supplementary data samples include static negative sampling data samples, dynamic negative sampling data samples and front-end click data samples; the distance values ​​between the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples and the incomplete valid data samples are calculated; the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples are sorted according to the increasing relationship of the distance values; the top K data samples with the smallest distance values ​​are selected, and the number of occurrences of each sample in the top K data samples is counted, and the integrity of the user data is greatly improved by supplementing the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples to the incomplete valid data samples.

[0037] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the claims.

[0038] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for analyzing user portraits, characterized in that: The method comprises: Obtaining an incomplete valid data sample, wherein the incomplete valid data sample is requested from a background; Acquire supplementary data samples, wherein the supplementary data samples include static negative sampling data samples, dynamic negative sampling data samples, and front-end click data samples; Calculating the distance value between the static negative sampling data sample, the dynamic negative sampling data sample, the front-end click data sample and the incomplete valid data sample; Sorting the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples according to the increasing relationship of the distance values; The first K data samples with the smallest distance values ​​are selected, and the number of occurrences of each sample in the first K data samples is counted.

2. The method for analyzing user portraits according to claim 1, characterized in that: If the distance value is greater than a preset value, the supplementary data sample is discarded. The distance value is specifically calculated by the following formula: ， in, and are the coordinates of the first and second sample points, is the distance value between the first sample point and the second sample point, the first sample point is extracted from the valid data sample, and the second sample point is extracted from the static negative sampling data sample, the dynamic negative sampling data sample, and the front-end click data sample.

3. The method for analyzing user portraits according to claim 1, characterized in that: The supplementary data sample is data within an activity cycle of the user. According to the time axis, the time from when the user enters the system to when the user exits the system is regarded as an activity cycle.

4. The method for analyzing user portraits according to claim 3, characterized in that: The time when the user enters the system and changes the token is considered an activity cycle.

5. The method for analyzing user portraits according to claim 3, characterized in that: The user's usage cycle on the system is regarded as an activity cycle.

6. The method for analyzing user portraits according to claim 1, characterized in that: The static negative sampling data samples are selected based on a preset distribution rule; the dynamic negative sampling data samples are selected based on the performance of the current model and the characteristics of the training data.

7. The method for analyzing user portraits according to claim 1, characterized in that: The more times each sample appears, the higher the confidence of the corresponding supplementary data sample.

8. A user portrait analysis device, characterized in that: The device comprises: A valid data sample acquisition unit, which is used to acquire an incomplete valid data sample, wherein the incomplete valid data sample is requested from a background; A supplementary data sample acquisition unit, which is used to acquire supplementary data samples, wherein the supplementary data samples include static negative sampling data samples, dynamic negative sampling data samples and front-end click data samples; A distance value calculation unit, which is used to calculate the distance value between the static negative sampling data sample, the dynamic negative sampling data sample, the front-end click data sample and the incomplete valid data sample; A data sample sorting unit, configured to sort the static negative sampling data samples, the dynamic negative sampling data samples and the front-end click data samples according to an increasing relationship of the distance values; A data sample statistical unit is used to select the first K data samples with the smallest distance value and count the number of occurrences of each sample in the first K data samples.

9. An electronic device, characterized in that: include: processor; as well as, A memory, configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images