Product list page reproduction method and device, storage medium and electronic device

Abstract

The present invention provides a method and device for reproducing a product list page, a storage medium and an electronic device, the method comprising: obtaining a complaint picture collected when a user complains about a problem in displaying a product list page of an application software, and obtaining a product information set after processing the complaint picture; based on the product information set, determining a target buried point record from multiple buried point records; obtaining an associated identifier in the target buried point record, obtaining user features, product features and a uniform resource locator associated with the associated identifier; based on the user features, product features and uniform resource locator, reproducing the product list page when the complaint picture was generated. The present invention quickly and accurately reproduces the product list page when the complaint picture was generated by using the user features, product features and uniform resource locator associated with the associated identifier, and provides developers with reliable data for analyzing problems.

Description

Technical Field

[0001] The present invention relates to the technical field of e-commerce, and in particular to a method and device for reproducing a commodity list page, a storage medium and an electronic device. Background Art

[0002] In the field of e-commerce, search and recommendation are two important means of shopping, and the user traffic brought by these two product forms generally accounts for more than 50%. If you provide users with a good experience and encourage them to place orders, you can increase the conversion rate and user value of user visits, and ultimately help e-commerce platforms increase the value of transaction goods; on the contrary, if the results of the product pages returned by search and recommendation affect the user experience, it may reduce the conversion rate and attract user complaints.

[0003] In order to provide users with a good experience, it is necessary to reproduce the product list page that caused user complaints, so that the R&D team can analyze the reproduced product display page and optimize the recommendation algorithm and search algorithm to provide users with high-quality services. Traditional reproduction of product list pages is usually based on screenshots and user identities provided by users, but traditional reproduction methods are difficult to accurately reproduce the product list page that caused user complaints. Summary of the invention

[0004] In view of this, the present invention provides a method and device for reproducing a product list page, a storage medium and an electronic device, which can obtain the URL, user characteristics and product characteristics through an associated identifier, and accurately reproduce the product list page that causes user complaints based on the acquired URL, user characteristics and product characteristics.

[0005] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0006] The first aspect of the present invention discloses a method for reproducing a product list page, comprising:

[0007] Obtaining complaint images, where the complaint images are images collected when users complain about problems with the product list page displayed by the application software;

[0008] Processing the complaint image to obtain a product information set, wherein the product information set includes product information of each product in the complaint image;

[0009] Based on the product information set, determine a target tracking record from a plurality of pre-generated tracking records, where the target tracking record is a record generated in the process of displaying the product list page to the user;

[0010] Obtaining an associated identifier in the target tracking record, and obtaining user features, product features, and a uniform resource locator associated with the associated identifier in a preset historical snapshot database;

[0011] Based on the user characteristics, the product characteristics and the uniform resource locator, the product list page when the complaint image was generated is reproduced.

[0012] In the above method, optionally, the step of processing the complaint image to obtain a product information set includes:

[0013] Using an image recognition algorithm to identify each product in the complaint image;

[0014] Determining the commodity information of each of the commodities in a preset commodity system;

[0015] According to the ranking position of each product in the complaint image, the product information of each product is sorted to generate a product information set.

[0016] In the above method, optionally, the step of determining a target tracking record from a plurality of pre-generated tracking records based on the product information set includes:

[0017] Determine the product set for each tracking record;

[0018] Comparing the product information set with each of the product sets;

[0019] Determine a product set including the product information set as a target product set;

[0020] The burying point record to which the target commodity set belongs is determined as the target burying point record.

[0021] In the above method, optionally, the process of pre-generating the target tracking point record includes:

[0022] Receive product browsing requests sent by users;

[0023] Based on the product browsing request, obtaining user information, product information and a uniform resource locator, wherein the uniform resource locator includes an associated identifier;

[0024] Acquire user characteristics corresponding to the user information and commodity characteristics corresponding to the commodity information;

[0025] Associating the user feature and the product feature with the association identifier;

[0026] Based on the user characteristics and the product characteristics, determining a product list page corresponding to the product browsing request;

[0027] The product list page corresponding to the product browsing request is displayed to the user, and whether a preset event tracking point is triggered in the process of displaying the product list page to the user is monitored in real time. When the event tracking point is detected to be triggered, the target tracking point record is generated based on the association identifier.

[0028] In the above method, optionally, determining the product list page corresponding to the product browsing request based on the user characteristics and the product characteristics includes:

[0029] Inputting the user characteristics and the product characteristics into a pre-trained product prediction model;

[0030] triggering the commodity prediction model to process the user features and the commodity features to obtain a commodity prediction list output by the commodity prediction model;

[0031] Based on the predicted commodity list, a commodity list page corresponding to the commodity browsing request is generated.

[0032] A second aspect of the present invention discloses a device for reproducing a product list page, comprising:

[0033] A first acquisition unit is used to acquire a complaint image, wherein the complaint image is an image collected when a user complains about a product list page displayed by the application software;

[0034] A processing unit, configured to process the complaint image to obtain a product information set, wherein the product information set includes product information of each product in the complaint image;

[0035] A first determining unit, configured to determine a target tracking record from a plurality of pre-generated tracking records based on the product information set, wherein the target tracking record is a record generated in the process of displaying the product list page to the user;

[0036] A second acquisition unit is used to acquire an associated identifier in the target tracking record, and acquire user features, product features and a uniform resource locator associated with the associated identifier in a preset historical snapshot database;

[0037] A reproduction unit is used to reproduce the product list page when the complaint image is generated based on the user characteristics, the product characteristics and the uniform resource locator.

[0038] In the above device, optionally, the processing unit includes:

[0039] An identification subunit, configured to identify each product in the complaint image using an image recognition algorithm;

[0040] A first determining subunit, used to determine the commodity information of each of the commodities in a preset commodity system;

[0041] The first generating sub-unit is used to sort the product information of each product according to the sorting position of each product in the complaint picture to generate a product information set.

[0042] In the above device, optionally, the first determining unit includes:

[0043] The second determination subunit is used to determine the commodity set of each tracking point record;

[0044] A comparison subunit, used for comparing the commodity information set with each of the commodity sets;

[0045] A third determining subunit, configured to determine a commodity set including the commodity information set as a target commodity set;

[0046] The fourth determining subunit is used to determine the burying point record to which the target commodity set belongs as the target burying point record.

[0047] The above device may optionally further include:

[0048] A receiving unit, used to receive a product browsing request sent by a user;

[0049] A third acquisition unit, configured to acquire user information, product information and a uniform resource locator based on the product browsing request, wherein the uniform resource locator includes an associated identifier;

[0050] a fourth acquisition unit, configured to acquire user characteristics corresponding to the user information and commodity characteristics corresponding to the commodity information;

[0051] an associating unit, configured to associate the user feature and the commodity feature with the association identifier;

[0052] A second determining unit, configured to determine a product list page corresponding to the product browsing request based on the user characteristics and the product characteristics;

[0053] A generation unit is used to display the product list page corresponding to the product browsing request to the user, and to monitor in real time whether a preset event tracking point is triggered in the process of displaying the product list page to the user. When the event tracking point is detected to be triggered, the target tracking point record is generated based on the association identifier.

[0054] In the above device, optionally, the second determining unit includes:

[0055] An input subunit, used for inputting the user features and the product features into a pre-trained product prediction model;

[0056] A triggering subunit, used to trigger the commodity prediction model to process the user features and the commodity features to obtain a commodity prediction list output by the commodity prediction model;

[0057] The second generating subunit is used to generate a product list page corresponding to the product browsing request based on the product prediction list.

[0058] A third aspect of the present invention discloses a storage medium, characterized in that the storage medium includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to execute the method for reproducing the product list page as described above.

[0059] A fourth aspect of the present invention discloses an electronic device, characterized in that it includes a memory and one or more instructions, wherein the one or more instructions are stored in the memory and are configured to be executed by one or more processors to perform the method for reproducing the product list page as described above.

[0060] Compared with the prior art, the present invention has the following advantages:

[0061] The present invention provides a method and device for reproducing a product list page, a storage medium and an electronic device, the method comprising: obtaining a complaint picture collected when a user complains about a problem in displaying a product list page of an application software, and obtaining a product information set after processing the complaint picture; based on the product information set, determining a target buried point record from multiple buried point records; obtaining an associated identifier in the target buried point record, obtaining user characteristics, product characteristics and a uniform resource locator associated with the associated identifier; based on the user characteristics, the product characteristics and the uniform resource locator, reproducing the product list page when the complaint picture was generated. The present invention accurately reproduces the product list page when the complaint picture was generated by using the user characteristics, product characteristics and uniform resource locator associated with the associated identifier, and provides developers with reliable data for analyzing problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0063] Figure 1 A flow chart of a method for displaying a page for searching products to a user;

[0064] Figure 2 A flow chart of a method for displaying a page of recommended products to a user;

[0065] Figure 3 A method flow chart of a method for reproducing a product list page provided by an embodiment of the present invention;

[0066] Figure 4 A method flow chart of another method for reproducing a product list page provided by an embodiment of the present invention;

[0067] Figure 5 A method flow chart of a method for reproducing a product list page provided by an embodiment of the present invention;

[0068] Figure 6 A method flow chart of another method for reproducing a product list page provided by an embodiment of the present invention;

[0069] Figure 7 A method flow chart of yet another method for reproducing a product list page provided by an embodiment of the present invention;

[0070] Figure 8 A scene example diagram of a method for reproducing a product list page provided by an embodiment of the present invention;

[0071] Fig. 9 A schematic diagram of the structure of a device for reproducing a product list page provided by an embodiment of the present invention;

[0072] Fig.10 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0073] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0074] In this application, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0075] In the field of e-commerce, search and recommendation are two important means of shopping, and the user traffic brought by these two product forms is generally more than 50%. If you provide users with a good experience and encourage them to place orders, you can improve the user visit conversion rate (UCVR) and user value (UV), and ultimately help e-commerce platforms increase the transaction value of goods (GMV); on the contrary, if the results of the product pages returned by search and recommendation affect the user experience, it may reduce the conversion rate and attract user complaints, and even cause legal liability.

[0076] Examples of common user experience issues: Repeatedly recommending recently purchased products to users, or low matching between searched products and search terms, wasting space on the App page to display products and prolonging user decision-making time.

[0077] Therefore, when receiving complaints about search and recommendation user experience, the R&D team needs to locate and analyze the root cause as soon as possible, solve it properly, shorten the impact time and scope of the bad user experience, and optimize the algorithm to avoid recurrence. These are one of the key tasks of the R&D team to improve the user experience.

[0078] There are three common ways for users to complain:

[0079] a) The user collects complaint images of the search or recommendation results of unsatisfactory product pages, and uploads them to the customer service for direct feedback through the complaint portal of the e-commerce software.

[0080] b) More radical users will post complaint pictures of e-commerce software on social platforms such as forums and bulletin boards to express their dissatisfaction; if these users happen to be influential users, it may cause major fluctuations in public opinion and affect the reputation of the e-commerce platform.

[0081] c) In addition, some well-intentioned users may forward the complaint images to staff working on the e-commerce platform or to their relatives and friends as a flexible feedback issue.

[0082] From the perspective of user experience, complaints about recommendations mainly include: recommending re-purchased products, recommending similar products in groups, sensitive products, and inconsistent images and text. Complaints about searches mainly focus on search accuracy and the products found being irrelevant to the search terms.

[0083] The search and recommendation recall products of mainstream e-commerce platforms are now tailored to each individual. The main approach is to use user portraits, product portraits, search terms (only applicable to search), geographical information, time information, and user behavior context as input features of deep neural networks, and learn the product recall set with the highest conversion rate through deep learning, and sort it according to a certain algorithm. Therefore, to analyze user complaints, at least the user's unique identifier (hereinafter referred to as the user ID) must be obtained; in addition, the Uniform Resource Locator (URL) of the user's access to the search and recommendation service request also contains a lot of information that can help the production and research team reproduce the request results and analyze user complaints.

[0084] When we receive complaints about search and recommendation user experience, we usually query the historical request URLs containing the user ID provided by the user in the backend log system. The following is an explanation of an exemplary search request URL: "key=iphone" is the search term entered by the user; "userid=linda" is the user ID; "area_ids=1,72,2839,0" is the user's geographic location information, which involves the product inventory and delivery area.

[0085] The following is an explanation of the content included in an exemplary search request URL: key=iphone&scene=0&log_id=09582426debe49198bf3453e56966420&pvid=09582426debe49198bf3453e56966420&uuid=linda&qp_exclude=24&multi_suppliers=yes&shop_col=yes&area_ids=1,72,2839,0&page=1&pagesize=20.

[0086] The entire process of user search, request URL, model prediction, and product list returned to the user is as follows: Figure 1 As shown, Figure 1 This is a flow chart of a method for displaying a page for searching products to users. When a user searches on the search page or opens the product recommendation page, the App sends a URL request to the backend service, which includes the search keyword and user_id. The backend service obtains "user features" and "product features" based on this information as input for "model prediction" and outputs the product recall set with the highest conversion rate, namely the "product list".

[0087] In short, "user features" and "product features" are the inputs of "model prediction", and "product list" is the output of "model prediction". Users complained about the result of "product list", so in order to analyze and locate the problem, the most important thing is for the R&D team to get the two inputs of "user features" and "product features". In the search scenario, you can generally see the "product list" and "search term" (key=iphone) through the user's screenshot, but you can't get the user information "user ID" (user_id=linda); if the problem exists for most user IDs, you may be able to reproduce and locate the problem, otherwise, you can only reproduce and find the cause if you find the user ID.

[0088] The analysis of user experience complaints in recommendation scenarios is more complicated than that in search scenarios. Figure 2 As shown, Figure 2 This is a flowchart of a page showing recommended products to users. The recommendation scenario does not have a search term, but a recommendation position identifier (rec_id). Through this identifier, the recommendation system finds a pool of products that can be recommended. In addition, there is no rec_id in the user's screenshot, and the recommended position information can only be roughly guessed through the picture; another challenge is that without a search term, the recommendation results of different users vary greatly, so the existing technical means are relatively limited in reproducing and analyzing recommendation complaints.

[0089] The existing methods for reproducing user complaints and locating problems have the following obvious shortcomings for the above three complaint methods:

[0090] For complaints of the type "direct feedback through the complaint portal", the "user ID" can be obtained, and the URL visited by the user at the time can be found in the system log. Only a part of the user complaints can be reproduced and a basis for analysis can be provided. The reason is that the user row obtained through the user ID is only a current snapshot, not a historical snapshot at the time of the user's complaint. However, user behavior and product information change very quickly, and the algorithm model recalls and arranges products for users by keenly sensing these changes. Therefore, the R&D team can reproduce the problem only when the "user behavior" and "product characteristics" have not changed, or the changes have not changed the results of the predicted "product list"; otherwise, even if the "user ID" is obtained, the problem cannot be reproduced.

[0091] For the two types of complaints, "posting complaint images to social platforms" and "forwarding complaint images to friends and relatives working on e-commerce platforms", it is generally not convenient to directly ask for the user's user ID because it involves personal privacy; even in the former type of complaint, "posting complaint images to social platforms", asking for the user ID may intensify the user's resistance. In this case, if it is a search scenario, you can also find some analysis clues through the search terms contained in the screenshots of the user and try to reproduce the problem; if it is a recommendation scenario, it is very difficult to reproduce the problem.

[0092] In summary, the existing methods for reproducing and analyzing user experience complaint problems are uncertain when only complaint images and user IDs are available, but no snapshots of user historical behavior and product information are available. The analysis results may be inaccurate due to unclear inputs.

[0093] The present invention aims to provide a method for reproducing a product list page. Through the associated identifier in the tracking record, the user characteristics, product characteristics and uniform resource locator associated with the associated identifier are determined, and the product list page that causes user complaints can be accurately reproduced through the user characteristics, product characteristics and uniform resource locator, providing accurate data reference for R&D personnel.

[0094] The present invention can be used in many general or special computing device environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multi-processor devices, distributed computing environments including any of the above devices or devices, etc.

[0095] The embodiment of the present invention provides a method for reproducing a product list page, the execution subject of the method is a backend server, referring to Figure 3 , which is a method flow chart of a method for reproducing a product list page provided by the present invention, and is specifically described as follows:

[0096] S101. Obtain complaint pictures.

[0097] The backend server obtains the complaint pictures. There are many ways to obtain the complaint pictures, such as the three methods mentioned above. The first is to obtain them directly through the complaint entrance of the e-commerce platform; the second is to obtain the complaint pictures sent by users on various social platforms; the third is that the user transfers the existing complaint pictures to the staff of the e-commerce platform.

[0098] There are many ways for users to collect complaint pictures. For example, users can use a mobile phone with a camera function to take a picture of the product page with complaints, so as to obtain a complaint picture. Users can also use the screenshot function built into the smart device to directly take a screenshot of the product page with complaints displayed on the smart device, so as to obtain a complaint picture.

[0099] The complaint image is a picture collected when a user complains about a product list page displayed by an application software, and the application software is software that can interact with an e-commerce platform for data; preferably, the complaint image includes multiple products predicted by the e-commerce platform based on a product browsing request sent by the user.

[0100] S102: Process the complaint image to obtain a product information set.

[0101] In the method provided by the embodiment of the present invention, the product information set includes product information of each product in the complaint image.

[0102] Reference Figure 4 , which is a flow chart of a method for processing a complaint image to obtain a product information set provided by an embodiment of the present invention, and is specifically described as follows:

[0103] S201. Use an image recognition algorithm to identify each product in the complaint image.

[0104] The image recognition algorithm is a preset algorithm. By using the image recognition algorithm to process complaint images, each product in the complaint image can be identified.

[0105] S202: Determine the commodity information of each commodity in a preset commodity system.

[0106] The commodity system includes but is not limited to a commodity search system and a commodity pool, wherein the commodity search system and the commodity pool contain information of multiple commodities.

[0107] When determining the product information of each product in the preset product system, you can first determine the product information of each product in the product retrieval system based on the scene type to which the complaint picture belongs. When the scene type to which the complaint picture belongs is the search product type, the product information of each product can be determined in the product retrieval system; when the scene type to which the complaint picture belongs is the recommended product type, the product information of each product can be determined in the product pool.

[0108] Product information includes but is not limited to product ID, product name, product category and other information.

[0109] S203. Sort the product information of each product according to the sorting position of each product in the complaint image to generate a product information set.

[0110] The various products are distributed and displayed in the complaint picture according to a preset arrangement, for example, they can be arranged and displayed in a sequence.

[0111] By arranging the product information of each product according to its sorting position in the complaint image, the orderliness of each product information can be ensured, so that the sorting position of each product information in the generated product information set is the same as the sorting position of each product in the complaint image, ensuring the consistency and correlation between the extracted data and the data provided by the user, and improving the accuracy of subsequent reproduction of the product list page corresponding to the complaint image.

[0112] S103. Based on the product information set, determine the target tracking record from multiple pre-generated tracking records.

[0113] In the method provided by an embodiment of the present invention, the target embedding record is a record generated in the process of displaying the product list page to the user.

[0114] Reference Figure 5 , is a flow chart of a method for determining target embedding records based on a commodity information set provided by an embodiment of the present invention, and is specifically described as follows:

[0115] S301. Determine the product set for each tracking record.

[0116] Tracking records are saved in the Elastic Search retrieval service. The Elastic Search retrieval service user saves the tracking records generated for each user.

[0117] The product set of the tracking point record contains multiple product IDs and an associated identifier, which is used to find the data used when generating a product list page for the user. Furthermore, each product ID in the product set is presented in a sequence, that is, each product ID in the product set forms a product ID sequence.

[0118] S302: Compare the product information set with each product set.

[0119] The product information set is compared with each product set to determine the product set that contains the product information set from among the multiple product sets.

[0120] Preferably, the product information set includes various product information arranged in sequence, that is, the various product information in the product information set forms a product information sequence. It should be noted that the product ID sequence in the product set including the product information set contains a sequence segment corresponding to the product information sequence in the product information set.

[0121] S303: Determine a product set including a product information set as a target product set.

[0122] S304. Determine the burying point record to which the target commodity set belongs as the target burying point record.

[0123] The target tracking record is determined by the product information set obtained by processing the complaint image. Thus, the record generated in the process of displaying the product list page to the user can be found among many tracking records. The complaint image is a picture collected based on the product list page. Therefore, the target tracking record is associated with the complaint image.

[0124] The present invention can query the target tracking record corresponding to the complaint image by using the Elastic Search retrieval service, so as to reproduce the product list page displayed to the user through the found record later.

[0125] S104: Obtain an associated identifier in the target tracking record, and obtain user features, product features, and a uniform resource locator associated with the associated identifier in a preset historical snapshot database.

[0126] The target tracking record includes an associated identifier, and the user characteristics, product characteristics and uniform resource locator corresponding to the associated identifier are determined in the historical feature snapshot database. It should be noted that the user characteristics and product characteristics are feature data applied when generating a product list page displayed to the user, and the uniform resource locator is used to obtain feature data applied when generating a product list page displayed to the user. Furthermore, the uniform resource locator also includes page information of the page type to which the product list page displayed to the user belongs.

[0127] S105. Based on the user characteristics, product characteristics and uniform resource locator, reproduce the product list page when the complaint image is generated.

[0128] When reproducing the product list page when the complaint image was generated, a product prediction list of multiple products can be obtained based on user characteristics and product characteristics, and a product list page can be generated based on the page information contained in the uniform resource locator and the product prediction list, wherein the generated product list page is the reproduced product list page.

[0129] Page types are divided into search type and recommendation type. The product list page corresponding to the search type is the product search page obtained after the user enters the search keyword; the product list page corresponding to the recommendation type is the product recommendation page that the e-commerce platform actively recommends to the user. The product list page corresponding to the recommendation type can be generated without searching for keywords.

[0130] In the method provided by the embodiment of the present invention, a complaint picture is obtained, which is a picture collected when a user complains about a product list page displayed by an application software; the complaint picture is processed to obtain a product information set, wherein the product information set includes product information of each product in the complaint picture; based on the product information set, a target tracking record is determined from multiple tracking records, wherein the target tracking record is a record generated in the process of displaying a product list page to the user; an associated identifier in the target tracking record is obtained, and user characteristics, product characteristics and a uniform resource locator associated with the associated identifier are obtained; based on the user characteristics, the product characteristics and the uniform resource locator, the product list page when the complaint picture was generated is reproduced. The present invention obtains the target tracking record corresponding to the complaint image, and can determine the user characteristics, product characteristics and uniform resource locator used to generate the product list page displayed to the user according to the associated identifier in the target tracking record. Therefore, the product list page displayed to the user when the complaint image is generated can be accurately reproduced according to the user characteristics, product characteristics and uniform resource locator. The reproduced product category list page can provide reliable data basis for research and development personnel, so that the research and development personnel can accurately optimize the recommendation algorithm and search method, and provide users with better quality services.

[0131] Reference Figure 6 , which is a flow chart of a method for pre-generating a target buried point record provided by an embodiment of the present invention, wherein the target buried point record is a buried point record, and therefore, the generation process of the buried point record can also be parameterized. Figure 6 The process shown is Figure 6 The specific instructions are as follows:

[0132] S401: Receive a product browsing request sent by a user.

[0133] There are two types of product browsing requests: one is a product search request, and the other is a product recommendation request. Among them, the product search request includes but is not limited to the search keywords entered by the user; the product recommendation request is a request generated when the user clicks on the recommendation page, and the request does not contain search keywords.

[0134] Preferably, the product browsing request includes an initial uniform resource locator, and the initial uniform resource locator does not include an associated identifier.

[0135] S402: Based on the product browsing request, obtain user information, product information and a uniform resource locator, wherein the uniform resource locator includes an associated identifier.

[0136] It should be noted that the uniform resource locator can be obtained by inserting the associated identifier into the initial uniform resource locator, wherein the associated identifier is a unique identifier.

[0137] User information includes but is not limited to user identity information, name, etc.; when the product browsing request is a product search request, the product information is the search keyword; when the product browsing request is a product recommendation request, the product information is the product recommendation identifier.

[0138] S403: Acquire user characteristics corresponding to the user information and commodity characteristics corresponding to the commodity information.

[0139] The user characteristics corresponding to the user information are searched in a preset user characteristic database, wherein the user characteristics include but are not limited to the behavior information of the goods recently browsed by the user, the information of the goods recently purchased, etc.

[0140] When the product information is a search keyword, the search keyword is processed using the product retrieval system to obtain product features corresponding to the product information; when the product information is a product recommendation identifier, the product features corresponding to the product recommendation identifier are extracted from the product recommendation pool; it should be noted that the product features include but are not limited to information such as product category and product production location.

[0141] S404: Associating the user characteristics and the product characteristics with the association identifier and saving them in a historical snapshot database.

[0142] The historical snapshot database may be composed of a kafka message system, which associates user features and product features with association identifiers, and saves the user features and product features associated with the association identifiers into the historical snapshot database.

[0143] By associating user characteristics and product characteristics with associated identifiers, it is easier to find the user characteristics and product characteristics when a complaint occurs later.

[0144] It should be noted that the historical feature snapshot database is used to save user behavior historical snapshots and product information historical snapshots. The user features saved in the historical feature snapshot database can be understood as user behavior historical snapshots, and the product features saved in the historical feature snapshot database can be understood as product information historical snapshots.

[0145] S405: Determine a product list page corresponding to the product browsing request based on the user characteristics and product characteristics.

[0146] Reference Figure 7 , is a flow chart of a method for generating a product list page corresponding to a product browsing request provided by an embodiment of the present invention, and is specifically described as follows:

[0147] S501: Input user features and product features into a pre-trained product prediction model.

[0148] The product prediction model is used to predict the products to be displayed to the user, wherein the product prediction model can be constructed using a deep neural network.

[0149] S502: triggering a product prediction model to process user features and product features to obtain a product prediction list output by the product prediction model.

[0150] After the user features and the product features are input into the product prediction model, the product prediction model processes the user features and the product features and outputs a prediction result, wherein the prediction result includes a product prediction list, and the product prediction list includes each predicted product.

[0151] S503: Generate a product list page corresponding to the product browsing request based on the product prediction list.

[0152] The embodiment of the present invention processes user features and product features by using a product prediction model to obtain a product prediction model, and generates a product list page according to the product prediction model.

[0153] S406. Display the product list page corresponding to the product browsing request to the user, and monitor in real time whether the preset event tracking point is triggered in the process of displaying the product list page to the user. When the triggering event tracking point is monitored, the target tracking point record is generated based on the association identifier.

[0154] It should be noted that there are many situations in which preset event tracking points are triggered. For example, when the product list page is displayed to the user, when the screenshot function of the smart terminal application is detected to display the product list page, the event tracking point is determined to be triggered. It should be noted that when the screenshot function of the smart terminal application is detected to display the product list page, the screenshot function of the smart terminal is used to collect complaint pictures by default when the user has a complaint about the product list page. Another example is that when the product list page is displayed to the user, the event tracking point is triggered immediately; it should be noted that the event tracking point is triggered when the product list page is displayed to the user in order to cover the situation where the user uses other picture collection devices to collect complaint pictures. Therefore, the event tracking point can be triggered when the product list page is displayed to the user.

[0155] After the event tracking is triggered, a target tracking record associated with the product list page displayed to the user is generated based on the associated identifier.

[0156] In the method provided by the embodiment of the present invention, a uniform resource locator including an associated identifier is obtained, and the user characteristics and product characteristics are associated and saved using the associated identifier, and a product list page is generated using the user characteristics and product characteristics. When an event tracking point is triggered in the process of displaying the product list page to the user, a tracking point record can be generated, so that when a complaint occurs later, the uniform resource locator, user characteristics and product characteristics can be determined based on the tracking point record, and the product list page displayed to the user can be reproduced based on the uniform resource locator, user characteristics and product characteristics, thereby providing R&D personnel with the data needed to resolve user complaints.

[0157] Reference Figure 8 , is a scene example diagram provided by an embodiment of the present invention, in which a scene of a user searching for a product is used for illustration, and the specific description is as follows:

[0158] The "search?key=iphone&user_id=linda" in the figure can be understood as the programming form of the product browsing request sent by the user, "ser_id=linda" can be understood as user information, and "key=iphone" can be understood as product information; the URL in the figure is a uniform resource locator containing an associated identifier; the user characteristics are obtained based on the user information, and the product retrieval system processes the product information to obtain product characteristics, and the user characteristics and product characteristics are input into the product prediction model. The product prediction model processes the user characteristics and product characteristics and outputs the prediction results. The prediction results include a product list, and a product list page is generated based on the product list in the prediction results, and the product list page is fed back to the user. In the process of displaying the product list to the user, when the user takes a screenshot, the event tracking is triggered, and the uniform resource locator, user characteristics, product characteristics and product list are associated using the associated identifier to generate a tracking record corresponding to the product list page displayed to the user. After receiving the screenshot photo provided by the user, the complaint image is identified and processed to obtain a product list (the product list at this time is equivalent to a product information set), and the corresponding tracking record is determined based on the product list. Based on the associated identifier in the tracking record, the user characteristics, product characteristics and uniform resource locator associated with the associated identifier are determined, and the product list page with the screenshot photo at the time of collection is reproduced based on the user characteristics, product characteristics and uniform resource locator.

[0159] In the method provided by the embodiment of the present invention, after obtaining the URL, the URL can be stored in the real-time log stream; a unique identifier page_id (equivalent to an associated identifier) ​​is added to the product list page displayed to the user and stored in the real-time log stream; the user uses a screenshot to trigger an event tracking, generates a tracking record, and stores it in the real-time log stream; wherein the tracking record is a record obtained by associating the URL with information related to the product list page using an associated identifier, and the record is a user screenshot tracking record that associates user characteristics, product characteristics, product list and URL.

[0160] When the R&D team obtains screenshots of user complaints (equivalent to complaint images), they can obtain a list of product IDs (equivalent to a collection of product information) based on the product retrieval system through manual or image recognition technology. This can be linked to the records of information related to the screenshots that have been stored (target tracking records). This information is sufficient to help the R&D team reproduce quickly and accurately, and provide clues for analyzing the problem.

[0161] Preferably, the solution provided by the present invention can be divided into three parts: storing information related to the product page; App triggering event tracking, and automatically associating information related to the product page in the background; identifying the product ID in the screenshot image, and searching for related feature information.

[0162] When a user uses an e-commerce app to search for products, or opens a push page to browse products (equivalent to receiving a product browsing request sent by the user), the backend system automatically "stores information related to the product page", which specifically includes the following steps:

[0163] The URL of the search page or recommendation page requested by the app is stored in the Kafka publish / subscribe message system, and the page_id field (which is used to store the association identifier) ​​is added to the URL as a unique identifier.

[0164] Obtain user features from the user information storage system, obtain product features from the product retrieval system or product pool, store the user features and product features in the Kafka message system respectively, and use the page_id field as a unique identifier for association when storing user features and product features.

[0165] The product prediction model processes product features and user features, outputs a product list, and generates a product list page based on the product list to provide feedback to the user. The page_id field is used as the unique identifier of the product list page and stored in the Kafka message system.

[0166] When a user uses the screenshot function to capture a product list page, an event tracking point is triggered, and the backend automatically associates the relevant information of the product page:

[0167] The user uses the screenshot function to capture a picture of the product list page and stores the captured picture in the local storage of the smart device.

[0168] User screenshot events and the unique identifier page_id of the product list page are stored in the Kafka message system.

[0169] The page_id of the screenshot event is associated with the "URL", "user features", and "product features" stored in the previous part to form a tracking record and stored in the Elastic Search retrieval service for users to query when they file complaints.

[0170] When the R&D team receives a screenshot photo (equivalent to a complaint photo) sent by a user, they can use the product ID (equivalent to a product information collection) to find the record in the Elastic Search retrieval service (equivalent to the target tracking record) and obtain the "URL" to reproduce the problem; use the "user characteristics" and "product characteristics" parameters to analyze the problem; use the obtained URL, user characteristics, and product characteristics to reproduce the product list page corresponding to the screenshot. The specific steps are as follows:

[0171] Manually analyze the product image in the screenshot and obtain the product ID in combination with the product search system, or automatically obtain the product ID through image recognition technology;

[0172] Use the product ID list to search the Elastic Search retrieval service to obtain the relevant records of the corresponding product page (equivalent to the target tracking record).

[0173] After the event tracking point is triggered, the "URL", "user characteristics", and "product characteristics" are associated into a tracking point record and stored in the Elastic Search retrieval service for users to query when they make complaints. After receiving the complaint image, the complaint image is processed to obtain the product ID, and the corresponding tracking point record is queried in the Elastic Search retrieval service to obtain the corresponding "URL", "user characteristics", and "product characteristics"; the "URL" is used to reproduce the problem; the "user characteristics" and "product characteristics" model input parameters are used to analyze the problem; the product list page when the user collected the complaint image can be reproduced based on the URL, user characteristics, and product characteristics.

[0174] The product list page reproduction solution provided by the embodiment of the present invention can be applied to a variety of user complaint scenarios and can comprehensively cover the user complaint channels. No matter what channel the user uses to make a complaint, the user characteristics, URL, and product characteristics that trigger the user complaint can be obtained. The product list page that triggers the user complaint can be quickly and accurately reproduced through the obtained user characteristics, URL, and product characteristics, providing the development team with the data basis required for repair and improving the efficiency of repairing and resolving complaints.

[0175] and Figure 1 Corresponding to the method shown in FIG. 1 , an embodiment of the present invention further provides a device for reproducing a product list page. The device can be applied to a backend server. The structural diagram of the device is shown in FIG. Fig. 9 The specific instructions are as follows:

[0176] The first acquisition unit 601 is used to acquire a complaint image, where the complaint image is an image collected when a user complains about a product list page displayed by the application software;

[0177] The processing unit 602 is used to process the complaint image to obtain a product information set, wherein the product information set includes product information of each product in the complaint image;

[0178] A first determining unit 603 is configured to determine a target tracking record from among a plurality of pre-generated tracking records based on the product information set, wherein the target tracking record is a record generated in the process of displaying the product list page to the user;

[0179] The second acquisition unit 604 is used to acquire the association identifier in the target tracking record, and acquire the user characteristics, product characteristics and uniform resource locator associated with the association identifier in a preset historical snapshot database;

[0180] The reproduction unit 605 is used to reproduce the product list page when the complaint image is generated based on the user characteristics, the product characteristics and the uniform resource locator.

[0181] In the device provided by the embodiment of the present invention, a complaint picture is obtained, which is a picture collected when a user complains about a product list page displayed by an application software; the complaint picture is processed to obtain a product information set, wherein the product information set includes product information of each product in the complaint picture; based on the product information set, a target burying point record is determined from multiple burying point records, wherein the target burying point record is a record generated in the process of displaying a product list page to the user; an associated identifier in the target burying point record is obtained, and user characteristics, product characteristics and a uniform resource locator associated with the associated identifier are obtained; based on the user characteristics, the product characteristics and the uniform resource locator, the product list page when the complaint picture was generated is reproduced. The present invention obtains the target tracking record corresponding to the complaint image, and can determine the user characteristics, product characteristics and uniform resource locator used to generate the product list page displayed to the user according to the associated identifier in the target tracking record. Therefore, the product list page displayed to the user when the complaint image is generated can be accurately reproduced according to the user characteristics, product characteristics and uniform resource locator. The reproduced product category list page can provide reliable data basis for research and development personnel, so that the research and development personnel can accurately optimize the recommendation algorithm and search method, and provide users with better quality services.

[0182] In the apparatus provided by the embodiment of the present invention, the processing unit 602 may be configured as follows:

[0183] An identification subunit, configured to identify each product in the complaint image using an image recognition algorithm;

[0184] A first determining subunit, used to determine the commodity information of each of the commodities in a preset commodity system;

[0185] The first generating sub-unit is used to sort the product information of each product according to the sorting position of each product in the complaint picture to generate a product information set.

[0186] In the apparatus provided in the embodiment of the present invention, the first determining unit 603 may be configured as follows:

[0187] The second determination subunit is used to determine the commodity set of each tracking point record;

[0188] A comparison subunit, used for comparing the commodity information set with each of the commodity sets;

[0189] A third determining subunit, configured to determine a commodity set including the commodity information set as a target commodity set;

[0190] The fourth determining subunit is used to determine the burying point record to which the target commodity set belongs as the target burying point record.

[0191] The device provided in the embodiment of the present invention may also be configured as:

[0192] A receiving unit, used to receive a product browsing request sent by a user;

[0193] A third acquisition unit, configured to acquire user information, product information and a uniform resource locator based on the product browsing request, wherein the uniform resource locator includes an associated identifier;

[0194] a fourth acquisition unit, configured to acquire user characteristics corresponding to the user information and commodity characteristics corresponding to the commodity information;

[0195] an associating unit, configured to associate the user feature and the commodity feature with the association identifier;

[0196] A second determining unit, configured to determine a product list page corresponding to the product browsing request based on the user characteristics and the product characteristics;

[0197] A generation unit is used to display the product list page corresponding to the product browsing request to the user, and to monitor in real time whether a preset event tracking point is triggered in the process of displaying the product list page to the user. When the event tracking point is detected to be triggered, the target tracking point record is generated based on the association identifier.

[0198] In the apparatus provided by the embodiment of the present invention, the second determining unit may be configured as follows:

[0199] An input subunit, used for inputting the user features and the product features into a pre-trained product prediction model;

[0200] A triggering subunit, used to trigger the commodity prediction model to process the user features and the commodity features to obtain a commodity prediction list output by the commodity prediction model;

[0201] The second generating subunit is used to generate a product list page corresponding to the product browsing request based on the product prediction list.

[0202] An embodiment of the present invention further provides a storage medium, which includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to execute the above-mentioned method for reproducing the product list page.

[0203] The embodiment of the present invention further provides an electronic device, the structural diagram of which is shown in FIG. Fig.10As shown, it specifically includes a memory 701 and one or more instructions 702, wherein the one or more instructions 702 are stored in the memory 701 and are configured to be executed by one or more processors 703 to execute the one or more instructions 702 to perform the above-mentioned method of reproducing the product list page.

[0204] The specific implementation processes and derivative methods of the above-mentioned embodiments are all within the protection scope of the present invention.

[0205] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0206] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0207] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for reproducing a product list page, characterized in that: include: Obtaining complaint images, where the complaint images are images collected when users complain about problems with the product list page displayed by the application software; Processing the complaint image to obtain a product information set, wherein the product information set includes product information of each product in the complaint image; Based on the product information set, determine a target tracking record from a plurality of pre-generated tracking records, where the target tracking record is a record generated in the process of displaying the product list page to the user; Obtaining an associated identifier in the target tracking record, and obtaining user features, product features, and a uniform resource locator associated with the associated identifier in a preset historical snapshot database; Based on the user characteristics, the product characteristics and the uniform resource locator, reproduce the product list page when the complaint image was generated; The process of pre-generating target tracking records includes: Receive product browsing requests sent by users; Based on the product browsing request, obtaining user information, product information and a uniform resource locator, wherein the uniform resource locator includes an associated identifier; Acquire user characteristics corresponding to the user information and commodity characteristics corresponding to the commodity information; Associating the user feature and the product feature with the association identifier; Based on the user characteristics and the product characteristics, determining a product list page corresponding to the product browsing request; The product list page corresponding to the product browsing request is displayed to the user, and whether a preset event tracking point is triggered in the process of displaying the product list page to the user is monitored in real time. When the event tracking point is detected to be triggered, the target tracking point record is generated based on the association identifier.

2. The method according to claim 1, characterized in that The processing of the complaint image to obtain a product information set includes: Using an image recognition algorithm to identify each product in the complaint image; Determining the commodity information of each of the commodities in a preset commodity system; According to the ranking position of each product in the complaint image, the product information of each product is sorted to generate a product information set.

3. The method according to claim 1, characterized in that The step of determining a target tracking point record from a plurality of pre-generated tracking point records based on the commodity information set includes: Determine the product set for each tracking record; Comparing the product information set with each of the product sets; Determine a product set including the product information set as a target product set; The burying point record to which the target commodity set belongs is determined as the target burying point record.

4. The method according to claim 1, characterized in that: The determining, based on the user characteristics and the product characteristics, a product list page corresponding to the product browsing request includes: Inputting the user characteristics and the product characteristics into a pre-trained product prediction model; triggering the commodity prediction model to process the user features and the commodity features to obtain a commodity prediction list output by the commodity prediction model; Based on the commodity prediction list, a commodity list page corresponding to the commodity browsing request is generated.

5. A device for reproducing a product list page, characterized in that: include: A first acquisition unit is used to acquire a complaint image, wherein the complaint image is an image collected when a user complains about a product list page displayed by the application software; A processing unit, configured to process the complaint image to obtain a product information set, wherein the product information set includes product information of each product in the complaint image; A first determining unit, configured to determine a target tracking record from a plurality of pre-generated tracking records based on the product information set, wherein the target tracking record is a record generated in the process of displaying the product list page to the user; A second acquisition unit is used to acquire an associated identifier in the target tracking record, and acquire user features, product features and a uniform resource locator associated with the associated identifier in a preset historical snapshot database; A reproduction unit, used to reproduce the product list page when the complaint image is generated based on the user characteristics, the product characteristics and the uniform resource locator; A receiving unit, used to receive a product browsing request sent by a user; A third acquisition unit, configured to acquire user information, product information and a uniform resource locator based on the product browsing request, wherein the uniform resource locator includes an associated identifier; a fourth acquisition unit, configured to acquire user characteristics corresponding to the user information and commodity characteristics corresponding to the commodity information; an associating unit, configured to associate the user feature and the commodity feature with the association identifier; A second determining unit, configured to determine a product list page corresponding to the product browsing request based on the user characteristics and the product characteristics; A generation unit is used to display the product list page corresponding to the product browsing request to the user, and to monitor in real time whether a preset event tracking point is triggered in the process of displaying the product list page to the user. When the event tracking point is detected to be triggered, the target tracking point record is generated based on the association identifier.

6. The device according to claim 5, characterized in that The processing unit comprises: An identification subunit, configured to identify each product in the complaint image using an image recognition algorithm; A first determining subunit, used to determine the commodity information of each of the commodities in a preset commodity system; The first generating sub-unit is used to sort the product information of each product according to the sorting position of each product in the complaint picture to generate a product information set.

7. The device according to claim 5, characterized in that The first determining unit includes: The second determination subunit is used to determine the commodity set of each tracking point record; A comparison subunit, used for comparing the commodity information set with each of the commodity sets; A third determining subunit, configured to determine a commodity set including the commodity information set as a target commodity set; The fourth determining subunit is used to determine the burying point record to which the target commodity set belongs as the target burying point record.

8. A storage medium, characterized in that: The storage medium includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to execute the method for reproducing a product list page according to any one of claims 1 to 4.

9. An electronic device, characterized in that: The system comprises a memory and one or more instructions, wherein the one or more instructions are stored in the memory and configured to be executed by one or more processors to perform the method for reproducing a product list page as claimed in any one of claims 1 to 4.

Images