Business Processing Method, Device, Equipment, Program Product and Storage Medium

Abstract

An embodiment of this specification provides a service processing method, apparatus, device, program product, and storage medium. The service processing method includes: determining that a target user logging in to a client belongs to a target user group, where the target user group includes users who have a preference for store evaluation behaviors determined in advance using historical store evaluation data of the user for the store party; displaying guiding information for guiding the target user to participate in store evaluation, where the guiding information is used to link to a store evaluation page; obtaining store evaluation data submitted by the target user through the store evaluation page after the guiding information is displayed, and identifying and displaying the store evaluation data that meets preset conditions.

Description

Technical Field

[0001] The embodiments of this specification relate to the field of Internet technologies, and particularly to methods, apparatuses, devices, program products, and storage media for business processing. Background Art

[0002] With the development of Internet technologies, more and more Internet service providers have set up servers, and business parties provide store clients for store parties. The store parties can display product information to the server through the store clients to publish the products in their stores. Business parties also provide clients for users. Users can purchase products from store parties through the clients. Among them, users can also evaluate the products of store parties. The evaluations of users play an important role for business parties and can better improve the services for users and stores. Based on this, it is necessary to optimize the user evaluation function to provide better services for users and store parties. Summary of the Invention

[0003] To overcome the problems existing in the related art, the embodiments of this specification provide methods, apparatuses, devices, program products, and storage media for business processing.

[0004] According to a first aspect of the embodiments of this specification, a business processing method is provided. The method includes:

[0005] Determine that the target user logging in to the client belongs to the target user group, where the target user group includes users who have a preference for store evaluation behaviors determined in advance by using the historical store evaluation data of the store party by the user;

[0006] Display guiding information for guiding the target user to participate in store evaluation, where the guiding information is used to link to the store evaluation page;

[0007] Obtain the store evaluation data submitted by the target user through the store evaluation page after the guiding information is displayed, and identify and display the store evaluation data that meets the preset conditions.

[0008] Optionally, the neural network includes a text recognition network, and the training data of the text recognition network includes: positive samples with a quality higher than the preset standard evaluation text calibrated in advance, and negative samples with a quality lower than the preset standard evaluation text calibrated in advance;

[0009] The text recognition network is used to: extract one or more of the following features of the evaluation text: features representing text richness, features representing text repetition, emotional features, semantic depth features, the correlation degree features between the text semantics and store products, or text layout features; and use the extracted features to predict whether the quality of the evaluation text is higher than the preset standard evaluation text.

[0010] Optionally, the features characterizing the text richness include one or more of the following: word count feature, number of words feature, number of phrases feature, or number of sentences feature;

[0011] And / or, the features characterizing the text repetition rate include one or more of the following: the feature of the word overlap rate in the text, the feature of the phrase overlap rate in the text, or the feature of the sentence overlap rate in the text.

[0012] Optionally, the neural network includes an image recognition network, and the image recognition network is used for:

[0013] After the image aesthetics sub-network extracts the features of the evaluation image, predicting the aesthetics score of the evaluation image; the training data of the image aesthetics sub-network includes: historical evaluation images whose user browsing information obtained from the historical evaluation data meets the set viewing rate condition;

[0014] And an image matching sub-network, which is used to predict the matching degree between the evaluation image and the product image of the store product;

[0015] Determine the quality of the evaluation image according to the aesthetics score, the matching degree, and the number of the evaluation images.

[0016] Optionally, after the step of identifying and displaying the store evaluation data that meets the preset conditions, the method further includes:

[0017] Obtain and display one or more of the following information: the access information of other users entering the store page corresponding to the displayed store evaluation data, the order conversion information of other users entering the store page corresponding to the displayed store evaluation data, and the change information of the store evaluation behavior of the target user.

[0018] According to the second aspect of the embodiments of the present specification, a service processing method is provided, and the method includes:

[0019] After detecting that the target user logs in, determine that the target user belongs to the target user group through the server, and the target user group includes: users who have a preference for the store evaluation behavior determined in advance by using the historical store evaluation data of the user for the store party;

[0020] Display the guiding information for guiding the target user to participate in the store evaluation, wherein the guiding information is used to link to the store evaluation page;

[0021] Obtain the store evaluation data submitted by the target user through the store evaluation page after displaying the guiding information, and send the store evaluation data to the server for the server to push after identifying the store evaluation data that meets the preset conditions.

[0022] According to a third aspect of the embodiments of the present specification, a service processing device is provided, and the device includes:

[0023] A determination module, configured to: determine that a target user logging in to the client belongs to a target user group, where the target user group includes: users who have a preference for store evaluation behavior determined in advance by using historical store evaluation data of the user for the store party;

[0024] A guidance module, configured to: display guidance information for guiding the target user to participate in store evaluation, where the guidance information is used to link to a store evaluation page;

[0025] A display module, configured to: obtain store evaluation data submitted by the target user through the store evaluation page after the guidance information is displayed, identify store evaluation data that meets preset conditions, and display it.

[0026] According to a fourth aspect of the embodiments of the present specification, a service processing device is provided, and the device includes:

[0027] A detection module, configured to: after detecting that a target user logs in, determine through the server that the target user belongs to a target user group, where the target user group includes: users who have a preference for store evaluation behavior determined in advance by using historical store evaluation data of the user for the store party;

[0028] A guidance module, configured to: display guidance information for guiding the target user to participate in store evaluation, where the guidance information is used to link to a store evaluation page;

[0029] A sending module, configured to: obtain store evaluation data submitted by the target user through the store evaluation page after the guidance information is displayed, and send the store evaluation data to the server for the server to identify store evaluation data that meets preset conditions and then push it.

[0030] According to a fifth aspect of the embodiments of the present specification, an electronic device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where when the processor executes the computer program, the steps of the method described in the first aspect or the second aspect are implemented.

[0031] According to a sixth aspect of the embodiments of the present specification, a computer storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect or the second aspect are implemented.

[0032] According to a seventh aspect of the embodiments of the present specification, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect or the second aspect are implemented.

[0033] The technical solutions provided by the embodiments of this specification may include the following beneficial effects:

[0034] In the embodiments of this specification, users with preferences for store evaluation behaviors can be determined in advance using the historical store evaluation data of the user for the store party. Therefore, these users may be able to provide more and higher-quality evaluation data. Based on this, after these target users log in to the client, guiding information for guiding the target users to participate in store evaluation is displayed to them. The guiding information is used to link to the store evaluation page, thereby reminding and guiding users with preferences for store evaluation behaviors to participate in store evaluation. The guiding information is used to link to the store evaluation page. Therefore, this guidance can not only meet the preferences of users but also remind users and facilitate users to conduct store evaluations. In addition, it is also designed to identify the evaluation data mentioned by these users, so as to screen out store evaluation data that meets the preset conditions and display it. Therefore, the above solutions can not only meet the needs of users who prefer store evaluation behaviors but also enable the business party to obtain evaluation data that meets the preset conditions, such as higher-quality evaluation data, etc. Through the display of these data, more information display opportunities are also brought to the store party.

[0035] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit this specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with this specification, and are used together with the specification to explain the principles of this specification.

[0037] Figure 1 FIG. is a schematic diagram of a business scenario shown according to an exemplary embodiment of this specification.

[0038] Figure 2A FIG. is a flowchart of a business processing method shown according to an exemplary embodiment of this specification.

[0039] Figure 2B FIG. is a schematic diagram of a page of a client shown according to an exemplary embodiment of this specification.

[0040] Figure 2C FIG. is a flowchart of another business processing method shown according to an exemplary embodiment of this specification.

[0041] Figure 3 FIG. is a hardware structure diagram of a computer device where a business processing device is located shown according to an exemplary embodiment of this specification.

[0042] Figure 4 It is a block diagram of a service processing device shown in accordance with an exemplary embodiment of this specification.

[0043] Figure 5 It is a block diagram of a service processing device shown in accordance with an exemplary embodiment of this specification. Detailed implementation manners

[0044] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with this specification. On the contrary, they are merely examples of devices and methods consistent with some aspects of this specification as detailed in the appended claims.

[0045] The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit this specification. The singular forms "a", "the", and "said" used in this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0046] It should be understood that although the terms first, second, third, etc. may be used in this specification to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this specification, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to determining".

[0047] As Figure 1 shown, it is a schematic diagram of a service scenario shown in accordance with an exemplary embodiment of this specification, including a service provider, a store, and a user. Generally, the service provider builds a server, provides a store client to the store, provides a user client to the user, and the store client and the user client are connected to the server. The store can use the functions provided by the store client to open a store, that is, a corresponding store page is configured for the store, and the store publishes products in the store; the user can use the user client to enter the store pages of each store, and can purchase the products provided by the store on the store page. Generally, the service provider also provides an evaluation function for the user to evaluate the store, such as an overall evaluation of the store or an evaluation of a certain product of the store. The evaluation content is displayed on the store page and can be viewed by each user who enters the store page.

[0048] Among them, the evaluations of users play an important role. For example, the evaluations displayed on the store page can be viewed by users, enabling them to understand the products of the store and helping them make decisions on whether to purchase the products. For the store, high-quality evaluations can attract users to purchase products and provide higher exposure rates for the store, etc.

[0049] Based on this, the business side usually provides an evaluation function on the order details page of the client after the product order between the user and the store is completed. Users can submit evaluation data through the evaluation function. Among them, the evaluation function allows users to submit one or more types of evaluation information. For example, it can include evaluations of different levels indicating the quality of the product, or it can allow users to submit image information, or it can allow users to submit text information, etc. In some other solutions, if the user does not actively submit evaluation data after a certain period of time after the product order is completed, the business side can also automatically generate evaluation data according to the set default evaluation method with the prior authorization and consent of the user, such as the default highest-level evaluation, etc.

[0050] Thus, it can be seen that how to obtain better evaluation data of users is of great significance for the business side to improve its business.

[0051] As Figure 2A shown, Figure 2A is a flowchart of a business processing method shown in this specification according to an exemplary embodiment. The method of this embodiment can be applied to the client or the server, and this method can include the following steps:

[0052] In step 202, determine that the target user logging in to the client belongs to the target user group, and the target user group includes: users who have a preference for store evaluation behaviors determined in advance by using the user's historical store evaluation data for the store.

[0053] In step 204, display guiding information for guiding the target user to participate in store evaluation, where the guiding information is used to link to the store evaluation page.

[0054] In step 206, obtain the store evaluation data submitted by the target user through the store evaluation page after the guiding information is displayed, and identify and display the store evaluation data that meets the preset conditions.

[0055] In this embodiment, in order to remind users to submit store evaluation data and to provide users with high-quality store evaluation data, in practical applications, different users have different preferences for store evaluation behaviors. In this embodiment, by designing a target user group, which indicates that users have preferences for store evaluation behaviors, only users belonging to the target user group are guided to participate in subsequent evaluation activities. In this embodiment, the target user group can be determined in advance by using the historical evaluation data of users on the store side.

[0056] In some examples, the historical evaluation of the store side by users is carried out when the users have a commodity order with the store side. The commodities in this embodiment can include labor products provided by the store side for users, which can be tangible items, or can also include intangible services such as housekeeping, hair cutting, or massage. The commodity information in this embodiment can include commodity names, commodity categories, commodity pictures, or commodity prices, etc. There can be various different commodity information according to the actual application scenario, and this embodiment does not make any limitations on this.

[0057] In some examples, with the consent of the user, historical evaluation data of each user on the store side can be obtained according to a preset time period, and various information related to evaluation behaviors of the user can be determined from the obtained historical evaluation data. As an example, the information related to evaluation behaviors can include the number of times of historical evaluation within the preset time period, for example, it can be the total number of times of historical evaluation within the preset time period, or the number of times of historical evaluation in each store within the preset time period; it can also include the frequency information of historical evaluation within the preset time period, for example, the ratio of the number of times of historical evaluation in each store within the preset time period to the number of stores, or the ratio of the total number of times of historical evaluation within the preset time period to the total number of all orders, etc.

[0058] In other examples, the historical evaluation data can also include the evaluation content of users; in practical applications, the evaluation content of some users may not be of high quality, for example, it is just a small amount of text, or the text submitted does not match the commodity information of the commodity order. Obviously, these low-quality evaluation data are not suitable for being pushed to users, and are of no help to users in understanding the store and its commodities, or in conducting commodity transactions. These low-quality evaluation data do not meet the expectations of users, and users do not want to spend time consulting on these low-quality evaluation data; based on this, this embodiment can also screen out historical evaluation data that meet preset conditions from the historical evaluation data.

[0059] In practical applications, the preset conditions can be flexibly configured according to needs. The preset conditions can include conditions indicating that the store evaluation data belongs to high-quality data, conditions indicating that the store evaluation data meets user expectations, conditions indicating that the store evaluation data contributes to users' understanding of store information, understanding of store product information, and / or conducting product transactions in the store, and so on. As an example, the preset conditions can be pre-configured by the service provider, which can include any combination of one or more rules. For example, it can include rules for the number of words in the evaluation text, rules for the semantics of the evaluation text, rules for the matching of the evaluation text with store information, rules for the matching of the evaluation text with store product information, rules for the number of evaluation images, rules for the content of evaluation images, rules for the aesthetics of evaluation images, rules for the matching of evaluation images with store information, or rules for the matching of evaluation images with store product information, and so on. Or, it can also be that the service provider prepares in advance evaluation texts with quality higher than the preset standard, and the preset conditions can include rules for matching with evaluation texts with quality higher than the preset standard; it can also be that the service provider prepares in advance evaluation images with quality higher than the preset standard, and the preset conditions can include rules for the matching of evaluation images with evaluation images with quality higher than the preset standard, and so on.

[0060] In some examples, there are various ways to select historical evaluation data that meets the preset conditions from historical evaluation data; as an example, historical evaluation data can be presented to the person screening historical evaluations through a human-computer interaction interface, and the selected historical evaluation data that meets the preset conditions can be obtained through the interaction interface. In other examples, historical evaluation data that meets the preset conditions can also be automatically screened through preset screening conditions, or a neural network can be pre-trained, and the historical evaluation data is input into the neural network, and the neural network predicts whether the historical evaluation data is historical evaluation data that meets the preset conditions.

[0061] Optionally, the historical store evaluation data includes evaluation texts and / or evaluation images, and the historical store evaluation data that meets the preset conditions is determined by a trained neural network for the evaluation texts and / or the evaluation images; among them, the implementation method of using the trained neural network for prediction will be provided with corresponding embodiments later.

[0062] In other examples, there can be other constraint conditions for determining the target user group. For example, it can be set to determine the target user group within a certain geographical range according to needs, and this embodiment does not limit this. Or, the target user group can have a certain timeliness, that is, the target user group is updated according to the set time period, and so on.

[0063] The determination process of the above-mentioned target user group can be pre - carried out on the server side. The server side can store each target user identifier in the target user group. In this embodiment, the user belonging to the target user group is called the target user. For step 202, if the server side executes the solution of this embodiment, when the client detects user login, it can obtain the user identifier of the logged - in user. After the server side obtains the user identifier of the logged - in user sent by the client, it can determine whether the logged - in user belongs to the target user group, that is, whether it is a target user. If the client executes the solution of this embodiment, the server side can send the determination result to the client, or only push the result that the logged - in user is a target user, and it can not push if the user is not a target user. Therefore, when the target user logs in to the client, the client can determine through the server side that it belongs to the target user group.

[0064] For users who do not belong to the target user group, this embodiment may not guide them to participate in the store evaluation. For target users, since they have a preference for the store evaluation behavior, this embodiment can execute step 204 to display the guiding information for guiding the target user to participate in the store evaluation, so as to remind and guide users with a preference for the store evaluation behavior to participate in the store evaluation. Among them, the guiding information is used to link to the store evaluation page. Therefore, this guidance can not only meet the user's preference, but also remind the user and facilitate the user to conduct the store evaluation. Among them, the store evaluation in this embodiment includes the user's evaluation of the purchased goods in the commodity order. Based on this, after determining that the user belongs to the target group, it can also be to execute step 204 after determining that the target user has a commodity order.

[0065] In some examples, there are various implementation methods for displaying the guiding information for guiding the target user to participate in the store evaluation; if this embodiment is executed by the server side, the server side can send a message to notify the client of the target user to display the guiding information through the client of the target user; among them, the guiding information can be pre - placed in the client, or pushed by the server side to the client. If this embodiment is executed by the client side, for example, the page displayed by the client of the target user includes the home page and multiple other non - home pages. This embodiment can display the guiding information on any page of the client, so that the user can view the guiding information when using the client. As an example, it can be displayed on one or more pages such as the home page or the commodity order page of the client.

[0066] In some examples, the guiding information is used to link to the store evaluation page, so that when the user views the guiding information, they can quickly trigger and enter the store evaluation page through the guiding information to conduct a store evaluation. Optionally, the guiding information can be implemented in various ways, such as including image information, text information, video information, or voice information, etc. This guiding information can be used to remind users to participate in the store evaluation and to remind users to trigger and directly link to the store evaluation page. As Figure 2B shown, it is a page schematic diagram of a client shown in this specification according to an exemplary embodiment. Figure 2B In it, a guiding information is shown in the client page of the target user, and this guiding information is exemplified by an image of a "loudspeaker" and the text "Write a high-quality evaluation and have a chance to win a 30-yuan red envelope".

[0067] After displaying the guiding information, the feedback data of the target user after displaying the guiding information can be obtained. In some examples, the target user may not trigger the guiding information to enter the store evaluation page. The feedback data in this embodiment can include the data recording that the target user did not trigger the guiding information to enter the store evaluation page. In other examples, the target user may also trigger the guiding information to enter the store evaluation page and conduct an evaluation. Therefore, the feedback data includes the store evaluation data submitted by the target user through the store evaluation page. Optionally, the process of obtaining the feedback data can last for a certain period of time, which can be flexibly determined according to needs in practical applications, and this embodiment does not limit this.

[0068] Among them, when the store evaluation data submitted by the target user through the store evaluation page is obtained, this embodiment can identify and display the store evaluation data that meets the preset conditions; among them, the implementation method of identifying the store evaluation data that meets the preset conditions can be the same as the implementation method of screening out the store evaluation data that meets the preset conditions from the historical evaluation data as described above. Optionally, the store evaluation data includes evaluation text and / or evaluation images; among them, the evaluation image can be an image uploaded by the user, or an image in a video uploaded by the user; the store evaluation data that meets the preset conditions is determined by a trained neural network for the evaluation text and / or the evaluation image; among them, the implementation method of using the trained neural network for prediction will be provided with corresponding embodiments later. Exemplarily, whether the store evaluation data submitted by the target user meets the preset conditions can be that the client sends the store evaluation data submitted by the user to the server, and the server executes the identification process. In other examples, the client can also pre-configure the preset conditions to screen out whether the store evaluation data submitted by the user meets the preset conditions, or pre-configure the aforementioned trained neural network, and the neural network conducts the identification, etc.

[0069] Exemplarily, the display of the store evaluation data that meets the preset conditions in this embodiment can be pushed by the server to the clients of each user for display. Optionally, in this embodiment, the identified store evaluation data that meets the preset conditions can be displayed using the page of the client; among them, the store evaluation or product evaluation is also displayed on the store page. Different from the store page, the page for displaying the store evaluation data in this embodiment can display all the identified store evaluation data that meets the preset conditions, that is, the page dedicated to displaying the store evaluation data can display the store evaluation data of multiple stores. As an example, this page can be the home page of the client, or a page at the next level of the home page of the client, etc.; for example, an entry to this page is provided on the home page of the client, and the user can jump to this page through this entry. Since this entry is on the home page of the client, the user can quickly enter this page after opening the client.

[0070] Optionally, in this embodiment, user rights and interests can also be granted to the users who identify the store evaluation data that meets the preset conditions. The user rights and interests in this embodiment refer to the incentives provided to the users to encourage the users to perform business behaviors. The rights and interests can be, for example, coupons, red envelopes, points, discount cards, special price products, virtual assets, etc., and can be flexibly configured according to the actual business needs.

[0071] In some examples, after the step of identifying and displaying store evaluation data that meets preset conditions, the method may further include: obtaining and displaying one or more of the following pieces of information: access information of other users entering the store page corresponding to the displayed store evaluation data, order conversion information of other users entering the store page corresponding to the displayed store evaluation data, and change information of the target user's store evaluation behavior. The steps in this embodiment can be executed by the server. By obtaining the above information, it can be provided for the service provider to improve the service. Among them, the display of this information can be pushed by the server to a preset terminal for display. The user of the preset terminal can be a relevant person on the service provider side for relevant personnel to consult, providing help for the business optimization of the service provider. For example, the access information may include the number of accesses; the order conversion information may include the order conversion rate, that is, the ratio of the number of orders to the number of accesses; the change information of the target user's store evaluation behavior may include changes in the number of evaluations of the target user, changes in the high-quality store evaluation data of the target user, etc. Based on this, by obtaining this information, it can be determined whether the store evaluation behavior of the target user has been enhanced, such as whether the number of store evaluations has increased, etc., or it can be determined whether the store evaluation data of the target user has been enhanced, such as whether the quantity of high-quality store evaluation data has increased, etc.; or, it can also be determined whether the business of the store side has been enhanced; for example, after displaying the store evaluation data that meets the preset conditions, it can be determined whether it has brought more traffic and order conversions to the store side, etc., so as to optimize subsequent operations, such as determining whether to increase or decrease the display of guiding information to guide users to participate in store evaluations; or determining whether to increase or decrease the rights and interests granted to users; or determining that positive effects can be brought after granting rights and interests, such as the store evaluation behavior of the target user has been enhanced, the quantity of high-quality store evaluation data has increased, the number of store evaluations has increased, and it has also brought more traffic and order conversions to the store, etc.

[0072] Exemplarily, when determining the target user group, it is necessary to identify whether the store evaluation data submitted by the user in history meets the preset conditions. In implementation, it can be that the server obtains the existing historical data and then conducts the identification to determine the target user group. For the identification of whether the store evaluation data submitted by the target user meets the conditions, it can be that the client sends the store evaluation data submitted by the user to the server, and the server executes it. Next, an embodiment is provided to illustrate the implementation manner of automatically identifying store evaluation data that meets the preset conditions by using a neural network.

[0073] In this embodiment, the task of the neural network is to score the input store evaluation data to determine whether the store evaluation data meets the preset conditions. The training process of the neural network can be as follows: first, represent a model through modeling, then evaluate the model by constructing an objective function, and finally optimize the objective function according to the training data and the optimization method to adjust the model to meet the set conditions. The structure of the neural network in this embodiment can be selected according to needs in actual applications, such as a convolutional neural network or a recurrent neural network, etc., and this embodiment does not limit this.

[0074] In some examples, the historical store evaluation data includes evaluation text and / or evaluation images; the users with preferences for store evaluation behaviors include: users who have historically submitted store evaluation data that meets the preset conditions; the store evaluation data that meets the preset conditions is determined by the trained neural network for the evaluation text and / or the evaluation images.

[0075] Identifying historical store evaluation data that meets the preset conditions, and identifying whether the store evaluation data uploaded by the user meets the preset conditions can be achieved using the same trained neural network. Among them, the store evaluation uploaded by the user can include text, images or videos; among them, for one or more images uploaded by the user, part or all of the images can be selected for recognition by the neural network. For videos, the video can be input into the neural network, and the neural network can extract all images or extract part of the images from the video for prediction; or, part of the images can also be extracted through a preset extraction algorithm and then input into the neural network. Optionally, taking the extraction of part of the images as an example, the rules for extracting images can be preset, for example, it can be extracting images at preset time intervals, or extracting images every N frames; it can also be identifying part of the images from the video, for example, using an object recognition algorithm, etc. to identify images with set targets from the video, and the set targets can be flexibly determined according to the business scenario, such as store goods, etc., and can be flexibly determined according to needs in actual applications, and this embodiment does not limit this.

[0076] Exemplarily, a neural network can be set for the text, which is called a text recognition network in this embodiment, to identify whether the evaluation text of the user meets the preset conditions; a image recognition network can be set for the image to identify whether the evaluation image of the user meets the preset conditions.

[0077] In some examples, the neural network includes a text recognition network, and the training data of the text recognition network includes: positive samples of evaluation text with a quality higher than the preset standard that are pre-calibrated, and negative samples of evaluation text with a quality lower than the preset standard that are pre-calibrated;

[0078] The text recognition network is used to: extract one or more of the following features of the evaluation text: features characterizing text richness, features characterizing text repetition, sentiment features, semantic depth features, the correlation degree features between text semantics and store products, or text layout features; and use the extracted features to predict whether the quality of the evaluation text is higher than that of a preset standard evaluation text.

[0079] In this embodiment, sample data for training can be prepared in advance. The sample data can include positive samples and negative samples. Among them, in this embodiment, preset standard evaluation texts are used to design positive samples and negative samples. The preset standard evaluation text can be preset in advance. The preset standard evaluation text refers to an evaluation text that accurately describes a product and / or a store, but the quality of this evaluation text is not high and may not contribute much to helping users understand the store and its products and make product transaction decisions. The quality of a positive sample is higher than that of the preset standard evaluation text, that is, the positive sample accurately describes the evaluated product and / or store, and the positive sample has high quality and can meet user expectations. After a user reads this evaluation, it can contribute to the user's understanding of the store and its products and making product transaction decisions. The quality of a negative sample is lower than that of the preset standard evaluation text, that is, the negative sample does not accurately describe the evaluated product and / or store, and / or the readability of the negative sample is poor. Users do not want to spend time reading this evaluation, and it does not contribute to the user's understanding of the store and its products and making product transaction decisions.

[0080] In practical applications, a preset standard evaluation text, positive samples, and negative samples can be constructed in advance. Exemplarily, positive samples and negative samples can be obtained by manually selecting from historical evaluation data in advance, or can be manually constructed, etc. For example, positive samples can be obtained from historical evaluation data based on information such as user reading volume and user browsing duration, such as samples with a large user reading volume and / or a long user browsing duration.

[0081] Exemplarily, a certain preset standard evaluation text can be: The taste is very good, and the delivery guy is very punctual; the positive sample can be: This potato tastes so delicious, sour and spicy, very appetizing with rice, and the delivery is also very punctual, highly recommended; the negative sample can be: Delicious. A sample with the same quality as the preset standard evaluation text is: The taste is very nice, and the delivery is also very punctual.

[0082] After preparing the above sample data, the text recognition network can be trained using the sample data. In practical applications, the type, structure, or parameters of the text recognition network can be set according to needs, and this embodiment does not limit this.

[0083] Another aspect of the training process is to select appropriate features. The features designed in this embodiment include: features representing text richness, features representing text repetition, sentiment features, semantic depth features, the correlation degree features between text semantics and store products, or text layout features. In practical applications, it can be set that the neural network extracts one or more of the above features from the evaluation text, so that the neural network can accurately learn the characteristics of high-quality evaluation texts, and then the input evaluation text can be accurately recognized subsequently.

[0084] Exemplarily, the features representing text richness include one or more of the following: word count feature, word number feature, phrase number feature, or sentence number feature. Based on this, when the neural network is training, it can count one or more of the word count, word number, phrase number, or sentence number of the evaluation text and convert them into corresponding features for learning. The above features can represent the richness of the text content. The richer the evaluation text content is, the higher the quality of the evaluation text is, which can meet the preset conditions and be helpful to users.

[0085] Exemplarily, the features representing text repetition include one or more of the following: the feature of word overlap rate in the text, the feature of phrase overlap rate in the text, or the feature of sentence overlap rate in the text. The above features can represent the overlapping degree of some content in the text content. In practical applications, some users, in order to submit more words, some content in the evaluation text appears repeatedly. Based on this, through the above-designed features, when the neural network is training, it can extract one or more of the above features representing text overlap degree. If the features representing text repetition are more, the quality of the evaluation text may be lower and fail to meet the preset conditions.

[0086] Exemplarily, the sentiment feature of the evaluation text is used to describe the user's views or attitudes towards the evaluated product / store, etc. It can be set that the neural network extracts each word in the evaluation text and determines it by analyzing the semantics, part of speech, or context association of the words. It can also pre-configure a word set that meets the preset conditions, and the neural network can extract the sentiment feature of the evaluation text by identifying the sentiment features of each word in the preset word set.

[0087] Exemplarily, it may further include a semantic depth feature of the evaluation text, which represents the depth of the text semantics of the product / store evaluated by the user in the evaluation text. The semantic depth is positively correlated with the quality of the evaluation text. It may further include a correlation degree feature between the text semantics and the store products, which represents whether the text semantics are related to the store products. For example, information such as the name, brand or category of the store products can be input into the neural network so that the neural network can extract the correlation degree feature between the text semantics and the store products. The degree of correlation between the text semantics and the store products is positively correlated with the quality of the evaluation text. It may further include a text layout feature, which represents the layout information of the text. For example, whether the text content is in paragraphs or all in one paragraph, and whether there are appropriate punctuation marks, etc.

[0088] In some examples, the neural network includes an image recognition network, and the image recognition network is used for:

[0089] After the image aesthetics sub-network extracts the features of the evaluation image, predicting the aesthetics score of the evaluation image; the training data of the image aesthetics sub-network includes: historical evaluation images whose user viewing rate meets the set viewing rate condition obtained from the historical evaluation data;

[0090] and an image matching sub-network, which is used to predict the matching degree between the evaluation image and the product image of the store products;

[0091] Determining the quality of the evaluation image according to the aesthetics score, the matching degree and the number of the evaluation images.

[0092] In practical applications, the type, structure or parameters of the image recognition network can be set as needed, and this embodiment does not limit this. As an example, a convolutional neural network can be used, etc.

[0093] This embodiment determines the quality of the evaluation image from three dimensions: image aesthetics, the matching degree between the image uploaded by the user and the store product image, and the number of evaluation images. The quality of the evaluation image is positively correlated with the evaluation image meeting the user's expectations.

[0094] In this embodiment, the image recognition network is designed to include two sub-networks: an image aesthetics sub-network and an image matching sub-network, which are respectively used to perform different tasks on the evaluation image.

[0095] In this embodiment, the image aesthetics sub-network can predict the aesthetics score of the input evaluation image, and the aesthetics score of the image measures the visual attractiveness of the image to the user. Based on this, for the business scenario of this embodiment, the training data of the image aesthetics sub-network includes historical evaluation images whose user access rates meet the set access rate conditions obtained from historical evaluation data, where the user access rate can be determined based on the user's historical click-through rate and / or the user's historical browsing information. The set access rate condition can be that the access rate is greater than the set access rate threshold, etc. The browsing information can include the browsing duration of each user or the number of browsing times of multiple users, etc. Therefore, the historical click-through rate and / or the user's historical browsing information are positively correlated with the access rate of the image, and also positively correlated with the visual attractiveness of the image to the user. Thus, these high-quality historical evaluation images are used to train the image aesthetics sub-network, enabling the network to learn the characteristics of images that meet the preset conditions, and thus having the ability to predict the aesthetics score of the input image. Optionally, the training data of the image aesthetics sub-network can also include other images, which can be flexibly determined according to actual needs in practical applications, and this embodiment does not limit this. Optionally, the features of the image can include one or more of the following features: features describing the simplicity, clarity, color, contrast, or average brightness of the image, and can also include features such as depth of field, depth of field, the rule of thirds, regional contrast, etc. These can be flexibly configured according to actual needs in practical applications.

[0096] In this embodiment, the image matching sub-network can predict the matching degree between the evaluation image and the product image of the store product. The store page of the store party is used to display the products of the store, and usually the store party uploads the product images of each product. In practical applications, however, users may upload images that do not match the product to be evaluated. Based on this, it is possible to identify whether the evaluation image uploaded by the user matches the product image to be evaluated, and this matching result is related to whether the evaluation image uploaded by the user meets the preset conditions. For example, the matching result is positively correlated with the quality. If the matching result is worse, the quality of the evaluation image is worse, and the probability of not meeting the preset conditions is higher.

[0097] In this embodiment, the quality of the evaluation image is also determined from the dimension of the number of images. For example, the number of images is positively correlated with the quality of the evaluation image.

[0098] Based on this, according to the aesthetic score, the matching degree, and the number of evaluation images, the quality of the evaluation images can be determined. Optionally, the weights of these three dimensions can be set as needed, and a score representing the quality of the evaluation images can be calculated based on a pre-designed calculation method. Alternatively, rules for determining the quality of the evaluation images can also be set. For example, for the dimension of the matching degree, it has a higher priority. If the evaluation images uploaded by the user have a poor matching degree with the product images of the store products, then the evaluation images probably do not meet the preset conditions. Even if the aesthetic score of the images themselves may be high and the number of images is large, since they are not related to the products being evaluated, the quality of the evaluation images is low. If the evaluation images uploaded by the user have a high matching degree with the product images of the store products, the aesthetic score and the number of evaluation images can then be determined, and the quality of the evaluation images can be determined based on the weights of the three. In practical applications, it can be flexibly determined as needed, and this embodiment does not limit it.

[0099] In some examples, the text recognition network and the image recognition network can be trained and run independently, or they can be connected and used as sub-networks of the neural network. This embodiment does not limit this.

[0100] Based on this, through the above embodiments, after the text recognition network is trained using positive samples and negative samples, it can accurately identify whether the input evaluation text meets the preset conditions for the input evaluation text. Through training, the image recognition network can accurately identify whether the input evaluation images meet the preset conditions for the input evaluation images.

[0101] In some examples, it is also possible to comprehensively determine whether the store evaluation data meets the preset conditions based on the prediction results of the text recognition network for the evaluation files, and / or based on the prediction results of the image recognition network for the evaluation images. As an example, the weights of these three dimensions can be set as needed, and a score representing the quality of the store evaluation data can be calculated based on a pre-designed calculation method, and then it can be determined whether the store evaluation data meets the preset conditions. Among them, the text recognition network and the image recognition network trained in the above embodiments can be used to identify whether the historical evaluation data meets the preset conditions, and can also be used to identify the evaluation data uploaded by the user to determine whether to display it.

[0102] As can be seen from the above embodiments, users with a preference for store evaluation behavior can be determined in advance using the user's historical store evaluation data for the store party. Therefore, these users may be able to provide more and higher-quality evaluation data. Based on this, it is designed to display guiding information for guiding the target users to participate in store evaluation after they log in to the client. The guiding information is used to link to the store evaluation page, so as to remind and guide users with a preference for store evaluation behavior to participate in store evaluation. The guiding information is used to link to the store evaluation page. Therefore, this guidance can not only meet the user's preference, but also remind the user and facilitate the user to conduct store evaluation. In addition, it is also designed to identify the evaluation data mentioned by these users, so as to screen out the store evaluation data that meets the preset conditions and display it. Therefore, the above solution can not only meet the needs of these users who prefer store evaluation behavior, but also enable the business party to obtain more and higher-quality evaluation data. Through the display of these data, it also brings more information display opportunities to the store party.

[0103] As mentioned above, Figure 2A The business processing method can be executed by the client or the server. As Figure 2C shown, it is a flowchart of another business processing method shown in this specification according to an exemplary embodiment, including the following steps:

[0104] In step 212, after detecting that the target user logs in, it is determined by the server that the target user belongs to the target user group, and the target user group includes: users who are determined in advance using the user's historical store evaluation data for the store party and have a preference for store evaluation behavior;

[0105] In step 214, display guiding information for guiding the target user to participate in store evaluation, where the guiding information is used to link to the store evaluation page;

[0106] In step 216, obtain the store evaluation data submitted by the target user through the store evaluation page after the guiding information is displayed, and send the store evaluation data to the server for the server to identify the store evaluation data that meets the preset conditions and then push it.

[0107] This embodiment describes the execution process of the business processing method from the perspective of the client as the execution subject.

[0108] Among them, in step 212, the client is for user use. When the user uses the client, the client can obtain the user identifier and initiate a login request by communicating with the server. The server can obtain the user identifier carried in the login request. As mentioned above, the server can pre-determine a target user group, and the target user group includes the user identifiers of multiple target users. It can match the user identifier carried in the login request with the user identifiers of the target users, so as to determine whether the user logging in to the client belongs to the target user group. The server can send the determination result that the logged-in user belongs to the target user group to the client, so that the client knows that the current logged-in user belongs to the target user group, and then executes the subsequent steps.

[0109] In step 216, after determining that the current logged-in user belongs to the target user group, the client can obtain the guiding information and display it; optionally, the guiding information can be pre-configured by the client, or can be the guiding information sent by the server after the client requests the server.

[0110] In step 218, the client can obtain the store evaluation data submitted by the target user through the store evaluation page after displaying the guiding information, and send the store evaluation data to the server. Based on this, the server can identify the store evaluation data that meets the preset conditions, and then push it to the clients of each user for display.

[0111] The implementation process of the above embodiments can refer to the description of the foregoing embodiments, and will not be elaborated here.

[0112] Corresponding to the embodiments of the foregoing service processing method, this specification also provides embodiments of a service processing apparatus and a computer device to which it is applied.

[0113] The embodiments of the service processing apparatus in this specification can be applied to a computer device, such as a server or a terminal device. The apparatus embodiments can be implemented through software, or through hardware or a combination of software and hardware. Taking software implementation as an example, as a logically meaningful apparatus, it is formed by the processor in its corresponding service processing reading the computer program instructions in the non-volatile memory into the memory for operation. From the hardware level, as Figure 3 shown, it is a hardware structure diagram of the computer device where the service processing apparatus in this specification is located. In addition to Figure 3 the shown processor 310, memory 330, network interface 320, and non-volatile memory 340, the computer device where the service processing apparatus 331 is located in the embodiment usually includes other hardware according to the actual functions of the computer device, which will not be elaborated here.

[0114] As Figure 4 shown, Figure 4It is a block diagram of a service processing device shown in this specification according to an exemplary embodiment. The device includes:

[0115] A determination module 41, configured to: determine that a target user logging in to the client belongs to a target user group, where the target user group includes: users who have a preference for store evaluation behaviors determined in advance using historical store evaluation data of the user for the store party;

[0116] A guidance module 42, configured to: display guidance information for guiding the target user to participate in store evaluation, where the guidance information is used to link to a store evaluation page;

[0117] A display module 43, configured to: obtain store evaluation data submitted by the target user through the store evaluation page after the guidance information is displayed, identify store evaluation data that meets preset conditions, and display it.

[0118] In some examples, the store evaluation data includes evaluation text and / or evaluation images; the users who have a preference for store evaluation behaviors include: users who have historically submitted store evaluation data that meets preset conditions; the historical store evaluation data that meets preset conditions is determined by a trained neural network after recognizing the evaluation text and / or the evaluation images.

[0119] In some examples, the neural network includes a text recognition network, and the training data of the text recognition network includes: positive samples of evaluation text with a quality higher than a preset standard that are pre-calibrated, and negative samples of evaluation text with a quality lower than the preset standard that are pre-calibrated;

[0120] The text recognition network is configured to: extract one or more of the following features of the evaluation text: features representing text richness, features representing text repetition, emotional features, semantic depth features, and the correlation degree features between the text semantics and store products; and use the extracted features to predict whether the quality of the evaluation text is higher than the preset standard evaluation text.

[0121] In some examples, the features representing text richness include one or more of the following: the number of characters feature, the number of words feature, the number of phrases feature, or the number of sentences feature;

[0122] And / or, the features representing text repetition include one or more of the following: the feature of the word overlap rate in the text, the feature of the phrase overlap rate in the text, or the feature of the sentence overlap rate in the text.

[0123] In some examples, the neural network includes an image recognition network, and the image recognition network is configured to:

[0124] After extracting the features of the evaluation image by the image aesthetics sub-network, predict the aesthetics score of the evaluation image; the training data of the image aesthetics sub-network includes: historical evaluation images whose user browsing information obtained from historical evaluation data meets the set viewing rate condition;

[0125] And an image matching sub-network, configured to predict the matching degree between the evaluation image and the product image of the store product;

[0126] Determine the quality of the evaluation image according to the aesthetics score, the matching degree, and the number of the evaluation images.

[0127] In some examples, the device further includes an information acquisition module, configured to: after the display module identifies store evaluation data that meets preset conditions and displays it, acquire and display one or more of the following information: access information of other users entering the store page corresponding to the displayed store evaluation data through the displayed store evaluation data, order conversion information of other users entering the store page corresponding to the displayed store evaluation data through the displayed store evaluation data, and change information of the target user's store evaluation behavior.

[0128] As Figure 5 shown, Figure 5 is a block diagram of a service processing device shown in this specification according to an exemplary embodiment. The device includes:

[0129] A detection module 51, configured to: after detecting that a target user logs in, determine that the target user belongs to a target user group through a server, where the target user group includes: users who have a preference for store evaluation behavior determined in advance by using the target user's historical store evaluation data for the store party;

[0130] A display module 52, configured to: display guiding information for guiding the target user to participate in store evaluation, where the guiding information is used to link to a store evaluation page;

[0131] A sending module 53, configured to: acquire the store evaluation data submitted by the target user through the store evaluation page after the guiding information is displayed, and send the store evaluation data to the server for the server to push after identifying the store evaluation data that meets the preset conditions.

[0132] Correspondingly, an embodiment of this specification further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the foregoing service processing method embodiment are implemented.

[0133] Correspondingly, an embodiment of this specification further provides a computer storage medium storing a computer program, which, when executed by a processor, implements the steps of the foregoing embodiment of the service processing method.

[0134] Correspondingly, an embodiment of this specification further provides a computer program product including a computer program, which, when executed by a processor, implements the steps of the foregoing embodiment of the service processing method.

[0135] The implementation processes of the functions and effects of each module in the foregoing service processing device are specifically described in the implementation processes of the corresponding steps in the foregoing service processing method, and will not be elaborated herein.

[0136] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can be referred to the partial description of the method embodiment. The device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this specification. Those of ordinary skill in the art can understand and implement it without creative work.

[0137] The specific embodiments of this specification are described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in a different order than in the embodiments and still achieve the desired results. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0138] The step divisions of the above various methods are only for clear description. When implemented, they can be combined into one step or some steps can be split into multiple steps, as long as the same logical relationship is included, and they are all within the protection scope of this patent; adding insignificant modifications to the algorithm or process or introducing insignificant designs, but not changing the core design of the algorithm and process are all within the protection scope of this application.

[0139] Among them, the description of "specific examples", or "some examples", etc. means that the specific features, structures, materials or characteristics described in combination with the embodiments or examples are included in at least one embodiment or example of this specification. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiments or examples. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0140] Those skilled in the art will readily conceive of other embodiments of the present specification after considering the specification and practicing the invention claimed herein. This specification is intended to cover any variations, uses, or adaptations of the present specification, which follow the general principles of the present specification and include common general knowledge or conventional technical means in the technical field not claimed in the present application. The specification and examples are only regarded as exemplary, and the true scope and spirit of the present specification are pointed out by the following claims.

[0141] It should be understood that the present specification is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present specification is only limited by the appended claims.

[0142] The above are only the preferred embodiments of the present specification and are not intended to limit the present specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present specification shall be included within the scope of protection of the present specification.

Claims

1. A business processing method, the method comprising: For multiple historical store evaluation data, screening out high-quality historical store evaluation data that meet preset conditions based on the following identification methods: using a trained text recognition network to identify whether the evaluation text in the store evaluation data belongs to high-quality evaluation text and using a trained image recognition network to identify whether the evaluation image in the store evaluation data belongs to high-quality evaluation image; Wherein, the training samples of the text recognition network include positive samples with calibrated quality higher than the preset standard evaluation text and negative samples with calibrated quality lower than the preset standard evaluation text; The image recognition network is used for: after extracting the features of the evaluation image by the image aesthetics sub-network, predicting the aesthetics score of the evaluation image; the training data of the image aesthetics sub-network includes: historical evaluation images whose user browsing information obtained from historical store evaluation data meets the set viewing rate condition; and an image matching sub-network for predicting the matching degree between the evaluation image and the product image of the store product; determining the quality of the evaluation image according to the aesthetics score, the matching degree, and the number of the evaluation images; The preset conditions include: the store evaluation data contains high-quality evaluation text and high-quality evaluation image; Determining that each user who has submitted the high-quality historical store evaluation data and has a preference for the store evaluation behavior belongs to the target user group; When the target user logs in to the client, displaying guiding information for guiding the target user to participate in the store evaluation, wherein the guiding information is used to link to the store evaluation page; Obtaining the store evaluation data submitted by the target user through the store evaluation page after the guiding information is displayed, and screening out high-quality target store evaluation data that meet the preset conditions by using the identification method; Displaying each of the high-quality target store evaluation data on the page of the client, so that the user can access the corresponding target store page through the displayed high-quality target store evaluation data.

2. The method according to claim 1, wherein the text recognition network is used for: extracting one or more of the following features of the evaluation text: features representing text richness, features representing text repetition, emotional features, semantic depth features, the correlation degree features between the text semantics and the store product, or text layout features; and predicting whether the quality of the evaluation text is higher than the preset standard evaluation text by using the extracted features.

3. The method according to claim 2, wherein the features representing text richness include one or more of the following: word count feature, word number feature, phrase number feature, or sentence number feature; And / or, the features representing text repetition include one or more of the following: the feature of word overlap rate in the text, the feature of phrase overlap rate in the text, or the feature of sentence overlap rate in the text.

4. The method according to claim 1, after the step of displaying each of the high-quality target store evaluation data on the page of the client, the method further includes: Obtain and display one or more of the following pieces of information: access information of other users entering the store page corresponding to the displayed store evaluation data through the store evaluation data, order conversion information of other users entering the store page corresponding to the displayed store evaluation data through the store evaluation data, and change information of the store evaluation behavior of the target user.

5. A service processing method, the method comprising: After detecting that a target user logs in, determine that the target user belongs to a target user group; the target user group is pre-determined by the server in the following manner: for multiple historical store evaluation data, filter out high-quality historical store evaluation data that meets the preset conditions based on the following identification method: use a trained text recognition network to identify whether the evaluation text in the store evaluation data belongs to high-quality evaluation text and use a trained image recognition network to identify whether the evaluation image in the store evaluation data belongs to high-quality evaluation image; wherein, the training samples of the text recognition network include positive samples with calibrated quality higher than the preset standard evaluation text and negative samples with calibrated quality lower than the preset standard evaluation text; the image recognition network is used for: after extracting the features of the evaluation image by the image aesthetics sub-network, predicting the aesthetics score of the evaluation image; the training data of the image aesthetics sub-network includes: historical evaluation images whose user browsing information obtained from historical store evaluation data meets the set viewing rate condition; and an image matching sub-network, used to predict the matching degree between the evaluation image and the product image of the store product; determine the quality of the evaluation image according to the aesthetics score, the matching degree, and the number of the evaluation images; the preset conditions include: the store evaluation data contains high-quality evaluation text and high-quality evaluation image; determine that each user who has submitted the high-quality historical store evaluation data and has a preference for the store evaluation behavior belongs to the target user group; display guiding information for guiding the target user to participate in the store evaluation, wherein the guiding information is used to link to the store evaluation page; Obtain the store evaluation data submitted by the target user through the store evaluation page after the guiding information is displayed, and send the store evaluation data to the server for the server to use the identification method to filter out high-quality target store evaluation data that meets the preset conditions and then push it, so that each high-quality target store evaluation data is displayed in the page of the client, so that the user can access the corresponding target store page through the displayed high-quality target store evaluation data.

6. A service processing device, the device comprising: A screening module, configured to: for multiple historical store evaluation data, screen out high-quality historical store evaluation data that meet preset conditions based on the following identification method: use a trained text recognition network to identify whether the evaluation text in the store evaluation data belongs to high-quality evaluation text and use a trained image recognition network to identify whether the evaluation image in the store evaluation data belongs to high-quality evaluation image; wherein, the training samples of the text recognition network include positive samples with calibrated quality higher than the preset standard evaluation text and negative samples with calibrated quality lower than the preset standard evaluation text; the image recognition network is configured to: after extracting the features of the evaluation image by an image aesthetics sub-network, predict the aesthetics score of the evaluation image; the training data of the image aesthetics sub-network includes: historical evaluation images whose user browsing information obtained from historical store evaluation data meets the set viewing rate condition; and an image matching sub-network, configured to predict the matching degree between the evaluation image and the product image of the store product; determine the quality of the evaluation image according to the aesthetics score, the matching degree, and the number of the evaluation images; the preset conditions include: the store evaluation data contains high-quality evaluation text and high-quality evaluation image; A determination module, configured to: determine that each user who has submitted the high-quality historical store evaluation data and has a preference for the store evaluation behavior belongs to the target user group; A guidance module, configured to: when the target user logs in to the client, display guidance information for guiding the target user to participate in the store evaluation, wherein the guidance information is used to link to the store evaluation page; A display module, configured to: obtain the store evaluation data submitted by the target user through the store evaluation page after the guidance information is displayed, and screen out high-quality target store evaluation data that meet the preset conditions by using the identification method; display each of the high-quality target store evaluation data on the page of the client, so that the user can access the corresponding target store page through the displayed high-quality target store evaluation data.

7. A service processing device, the device includes: A detection module, configured to: after detecting the login of a target user, determine that the target user belongs to a target user group; the target user group is determined by the server in advance in the following manner: for multiple historical store evaluation data, high-quality historical store evaluation data that meets preset conditions is screened out based on the following identification method: whether the evaluation text in the store evaluation data belongs to high-quality evaluation text is identified by a trained text recognition network and whether the evaluation image in the store evaluation data belongs to high-quality evaluation image is identified by a trained image recognition network; wherein, the training samples of the text recognition network include positive samples with calibrated quality higher than the preset standard evaluation text and negative samples with calibrated quality lower than the preset standard evaluation text; the image recognition network is configured to: after the aesthetic feature sub-network extracts the features of the evaluation image, predict the aesthetic score of the evaluation image; the training data of the aesthetic feature sub-network of the image recognition network includes: historical evaluation images whose user browsing information obtained from historical store evaluation data meets the set viewing rate condition; and an image matching sub-network, configured to predict the matching degree between the evaluation image and the product image of the store product; determine the quality of the evaluation image according to the aesthetic score, the matching degree and the number of the evaluation images; the preset condition includes: the store evaluation data includes high-quality evaluation text and high-quality evaluation image; determine that each user who has submitted the high-quality historical store evaluation data and has a preference for the store evaluation behavior belongs to the target user group; A display module, configured to: display guiding information for guiding the target user to participate in the store evaluation, wherein the guiding information is used to link to the store evaluation page; An identification module, configured to: obtain the store evaluation data submitted by the target user through the store evaluation page after the guiding information is displayed, and send the store evaluation data to the server for the server to screen out high-quality target store evaluation data that meets the preset conditions by using the identification method and then push it, so that each piece of the high-quality target store evaluation data is displayed in the page of the client, so that the user can access the corresponding target store page through the displayed high-quality target store evaluation data.

8. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

9. A computer program product, including a computer program, when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

10. A computer-readable storage medium, on which a computer program is stored, when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images