Product information matching and publishing method and its device, equipment, medium, and product

Abstract

The present application discloses a method for matching and publishing product information and its device, equipment, medium, and product. The matching method includes: obtaining the title text of the product to be published submitted by the terminal device; extracting the deep semantic information of the title text to obtain the corresponding text feature vector; calculating the similarity between the text feature vector and the text feature vector of the detail text of the historical product in the product feature library, and sorting the detail text of the historical product according to the similarity; pushing the sorted detail text to the terminal device for display as a list of alternative detail texts for the product to be published. The present application facilitates the entry of product information of the product to be published, and can intelligently recommend the detail text of the product information based on the title text of the product information that has been entered. It can serve a variety of application scenarios including the e-commerce field and has wide versatility.

Description

Technical Field

[0001] The present application relates to the field of e-commerce information technology, and in particular to a product information matching and publishing method and its corresponding device, computer equipment, computer-readable storage medium, and computer program product. Background Art

[0002] In the e-commerce platform, there are many application scenarios that require the input of product information. For example, when a merchant user needs to publish a new product, the product information of the product to be published needs to be submitted; when a merchant user needs to publish a product advertisement, the product information of the product to be published also needs to be submitted. For these application scenarios, the user needs to access the corresponding input page of the e-commerce platform on a mobile terminal or a computer terminal, and enter the corresponding product information on the page so as to submit it to the server for product publishing.

[0003] The product information includes a product title for summarizing the comprehensive information of the product, product details for describing the detailed information of the product (including for the purpose of advertising), and product pictures for displaying the shape, structure, color, and other contents of the product. Among them, the title text corresponding to the product title and the detail text corresponding to the product details are both in the form of text. The title text is relatively simple, but the detail text is relatively complex and delicate. When the merchant needs to fill in the detail text, especially for mobile terminals, the input and editing process is very cumbersome and inefficient. Since there is no relevant auxiliary technology in the prior art, in the face of this situation, even if there is relevant product information for products similar to the products to be released, users cannot directly quote it. Therefore, it brings many inconveniences to users in operation and the efficiency of product information release is low.

[0004] On the server side, there is a huge amount of product information, and the data value of this product information needs to be mined. If the data value of this product information is not effectively mined, it will also cause information waste. Especially when users publish product information, if the server cannot improve the level of intelligent service, it will affect the operating efficiency of the e-commerce platform and its overall user experience.

[0005] In summary, it is necessary to conduct in-depth research on technologies related to the matching and release of product information in order to explore more technical solutions that are suitable for serving actual needs. Summary of the invention

[0006] The primary purpose of the present application is to solve at least one of the above problems and to provide a product information matching method and its corresponding device, computer equipment, computer-readable storage medium, and computer program product.

[0007] In order to meet the various objectives of this application, this application adopts the following technical solutions:

[0008] A product information matching method provided for one of the purposes of this application includes the following steps:

[0009] Obtain the title text of the product to be released submitted by the terminal device;

[0010] Extracting deep semantic information of the title text to obtain a corresponding text feature vector;

[0011] Calculating the similarity between the text feature vector and the text feature vector of the detail text of the historical product in the product feature library, and sorting the detail text of the historical product according to the similarity;

[0012] The sorted detail texts are pushed to the terminal device for display as a list of candidate detail texts for the products to be released.

[0013] In a further embodiment, calculating the similarity between the text feature vector and the text feature vector corresponding to the detail text of the historical commodity in the commodity feature library, and sorting the detail text of the historical commodity according to the similarity comprises the following steps:

[0014] Calculating the similarity between the text feature vector of the title text of the product to be released and the text feature vector of the detail text of the historical products in the product feature library to obtain a historical product similarity sequence, wherein each of the text feature vectors is extracted using the same pre-trained neural network model;

[0015] According to the similarity, the historical products in the historical product similarity sequence whose similarity exceeds a preset threshold are sorted to obtain a first recommendation group;

[0016] A detailed text recommendation list is constructed for being pushed to the terminal device for display, and the detailed text recommendation list includes the first recommendation group.

[0017] In an extended embodiment, after the step of sorting the historical commodities in the historical commodity similarity sequence whose similarity exceeds a preset threshold according to the similarity and obtaining the first recommendation group, the following steps are also included:

[0018] According to the advertisement data of the historical products, the historical products with similarities lower than a preset threshold in the historical product similarity sequence are sorted to obtain a second recommendation group, and the second recommendation group is connected after the first recommendation group in the detail text recommendation list.

[0019] In a specific embodiment, according to the advertisement data of historical products, historical products with similarity lower than a preset threshold in the historical product similarity sequence are sorted to obtain a second recommendation group, including the following steps:

[0020] Extract historical products whose similarity is lower than a preset threshold from the historical product similarity sequence to obtain a temporary calculation list;

[0021] Calling multiple pieces of advertising data of each historical product in the temporary calculation list, wherein the advertising data includes any one or multiple pieces of the corresponding historical product advertising time, advertising time, and advertising effectiveness data;

[0022] For each historical product in the temporary calculation list, weighted summing up multiple pieces of advertising data to obtain a corresponding temporary score;

[0023] The historical commodities in the temporary calculation list are sorted according to the temporary scores to obtain the second recommendation group.

[0024] A method for publishing product information provided to meet one of the purposes of this application includes the following steps:

[0025] Displaying a product information input interface for the product to be released, the input interface including a first input box corresponding to the title text in the product information and a second input box corresponding to the detail text;

[0026] In response to a preset trigger event acting on the second input box, submitting the title text in the first input box to the server to obtain a list of candidate detail texts semantically matching the title text, the list including detail texts corresponding to a plurality of historical products;

[0027] Outputting the candidate detail text list to the current input interface to display detail texts from a plurality of historical commodities, so that each detail text is added to the second input box in response to the touch operation;

[0028] In response to the information submission event acting on the input interface, the product information displayed in the first input box and the second input box is submitted to the server to execute the release of the product to be released.

[0029] In a preferred embodiment, the preset trigger event acting on the second input box is a text input event generated by inputting text in the first input box.

[0030] In an extended embodiment, after the step of outputting the candidate detail text list to the current input interface to display detail texts from a plurality of historical commodities so that each detail text is added to the second input box in response to a touch operation, the following steps are included:

[0031] In response to a selection event acting on any detail text in the candidate detail text list, the selected detail text is connected to the input content in the second input box and displayed.

[0032] A product information matching device provided to meet one of the purposes of the present application includes: a title acquisition module, a semantic extraction module, a similarity matching module, and a push display module, wherein the title acquisition module is used to obtain the title text of the product to be released submitted by the terminal device; the semantic extraction module is used to extract the deep semantic information of the title text to obtain the corresponding text feature vector; the similarity matching module is used to calculate the similarity between the text feature vector and the text feature vector of the detail text of the historical product in the product feature library, and sort the detail text of the historical product according to the similarity; the push display module is used to push the sorted detail text to the terminal device for display as a list of alternative detail texts for the product to be released.

[0033] In a further embodiment, the similarity calculation module includes: a similarity calculation submodule, which is used to calculate the similarity between the text feature vector of the title text of the product to be released and the text feature vector of the detail text of the historical product in the product feature library, and obtain a historical product similarity sequence, and each of the text feature vectors is extracted using the same pre-trained neural network model; a similarity selection submodule, which is used to sort the historical products in the historical product similarity sequence whose similarity exceeds a preset threshold according to the similarity, and obtain a first recommendation group; a list push submodule, which is used to construct a detail text recommendation list for pushing to the terminal device for display, and the detail text recommendation list includes the first recommendation group.

[0034] In an extended embodiment, the similarity calculation module also includes: a historical preference sub-module, which is used to sort the historical products in the historical product similarity sequence whose similarity is lower than a preset threshold according to the advertising data of the historical products, and obtain a second recommendation group, and the second recommendation group is connected after the first recommendation group in the detail text recommendation list.

[0035] In a specific embodiment, the historical optimization submodule includes: a low-value extraction unit, which is used to extract historical products with a similarity lower than a preset threshold from the historical product similarity sequence to obtain a temporary calculation list; a data calling unit, which is used to call multiple advertising data of each historical product in the temporary calculation list, and the advertising data includes any one or multiple of the corresponding historical product advertising duration, advertising duration, and advertising effectiveness data; a comprehensive quantification unit, which is used to weightedly sum multiple advertising data for each historical product in the temporary calculation list to obtain a corresponding temporary score; a sorting processing unit, which is used to sort the historical products in the temporary calculation list according to the temporary score to obtain the second recommendation group.

[0036] A product information publishing device provided to meet one of the purposes of the present application includes: an interface display module, an alternative acquisition module, a text display module, and a submission and publishing module, wherein the interface display module is used to display a product information entry interface of the product to be released, and the entry interface includes a first input box corresponding to the title text in the product information and a second input box corresponding to the detail text; the alternative acquisition module is used to respond to a preset trigger event acting on the second input box, submit the title text in the first input box to a server to obtain a list of alternative detail texts semantically matching the title text, and the list includes detail texts corresponding to multiple historical products; the text display module is used to output the list of alternative detail texts to the current entry interface to display detail texts from multiple historical products, so that each detail text is added to the second input box in response to a touch operation; the submission and publishing module is used to respond to the information submission event acting on the entry interface, and submit the product information displayed in the first input box and the second input box to the server to execute the release of the product to be released.

[0037] In a preferred embodiment, the preset trigger event acting on the second input box is a text input event generated by inputting text in the first input box.

[0038] In an extended embodiment, the product information publishing device also includes: a selection insertion module, which is used to respond to a selection event acting on any detail text in the alternative detail text list, and connect the selected detail text to the input content in the second input box and then display it.

[0039] A computer device provided to meet one of the purposes of the present application includes a central processing unit and a memory, wherein the central processing unit is used to call and run a computer program stored in the memory to execute the steps of the product information matching method or the product information publishing method described in the present application.

[0040] A computer-readable storage medium is provided to meet another purpose of the present application, which stores a computer program implemented according to the product information matching method or product information publishing method in the form of computer-readable instructions. When the computer program is called and executed by a computer, the steps included in the method are executed.

[0041] A computer program product provided to meet another purpose of the present application includes a computer program / instruction, which implements the steps of the method described in any embodiment of the present application when executed by a processor.

[0042] Compared with the prior art, the advantages of this application are as follows:

[0043] Firstly, the present application implements semantic matching on the server side according to the text feature vector of the title text in the product information of the product to be released, and calculates the similarity between the text feature vector and the text feature vector of the detail text in the product information of the historical product, and determines the detail texts of multiple historical products that are semantically associated with the title text of the product to be released according to the similarity, and pushes them to the terminal device as an optional option for entering the detail text in the product information of the product to be released, thereby improving the intelligence level of the server's product entry service, and further improving the product information entry efficiency on the user side, providing convenience for quick release in application scenarios such as product advertising and online release.

[0044] Secondly, the present application implements, on the terminal device side, obtaining a list of semantically related alternative detail texts from the server based on the title text entered by the user in the first input box corresponding to the product title of the entry interface of the product to be released, and using each detail text in the list as an optional option for the input content of the second input box of the entry interface for entering the detail text of the product to be released. As a result, the user only needs to enter the title text of the product title of the product to be released in the first input box, and can obtain optional detail text that is semantically similar to the title text in the second input box. The user can quickly modify and edit the second input box of the product to be released based on the selection of a detail text, thereby realizing fast and convenient input of product details of the product to be released, improving the editing efficiency of the product information of the product to be released, and obtaining a good user experience.

[0045] In addition, in the business logic implemented in this application, semantic matching is performed between the title text of the product to be released and the detail text of the historical product, rather than semantic matching between title texts. Therefore, the alternative detail text can be directly hit through the title text. Since the two are semantically associated, the matching degree is higher and the hit is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0047] Figure 1 A flowchart of a typical embodiment of the commodity information matching method of the present application;

[0048] Figure 2 A schematic diagram of a process for constructing the first recommendation group in an embodiment of the present application;

[0049] Figure 3 A schematic diagram of a process for constructing a detailed text recommendation list in an embodiment of the present application;

[0050] Figure 4A schematic diagram of a process for constructing the second recommendation group in an embodiment of the present application;

[0051] Figure 5 A flowchart of a typical embodiment of the method for publishing product information of the present application;

[0052] Figure 6 A schematic diagram of an alternative detail text list execution without a control display mechanism in a preferred embodiment of the present application;

[0053] Figure 7 This is a schematic diagram of the process of product release in the embodiment of this application;

[0054] Figure 8 This is a diagram of an interface for inputting product information of a product to be released in one embodiment of the present application;

[0055] Fig. 9 It is a display diagram of recommended copy in the product information entry interface of the product to be released in the embodiment of the present application;

[0056] Fig.10 This is a diagram of an interface for inputting product information of a product to be released in another embodiment of the present application;

[0057] Fig.11 This is a principle block diagram of the commodity information matching device of this application;

[0058] Fig.12 This is a functional block diagram of the product information publishing device of this application;

[0059] Fig.13 A schematic diagram of the structure of a computer device used in this application. DETAILED DESCRIPTION

[0060] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be interpreted as limiting the present application.

[0061] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0062] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with those in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as here.

[0063] It will be understood by those skilled in the art that the "client", "terminal" and "terminal device" used herein include both devices with wireless signal receivers, which are devices with only wireless signal receivers without transmission capabilities, and devices with receiving and transmitting hardware, which are devices with receiving and transmitting hardware capable of two-way communication on a two-way communication link. Such devices may include: cellular or other communication devices such as personal computers, tablet computers, which have single-line displays or multi-line displays or cellular or other communication devices without multi-line displays; PCS (Personal Communications Service, personal communication system), which can combine voice, data processing, fax and / or data communication capabilities; PDA (Personal Digital Assistant, personal digital assistant), which may include a radio frequency receiver, pager, Internet / intranet access, web browser, notepad, calendar and / or GPS (Global Positioning System, global positioning system) receiver; conventional laptop and / or palmtop computers or other devices, which have and / or include a conventional laptop and / or palmtop computer or other device with and / or including a radio frequency receiver. The "client", "terminal" and "terminal device" used herein may be portable, transportable, installed in a vehicle (air, sea and / or land), or suitable for and / or configured to run locally, and / or in a distributed form, at any other location on the earth and / or in space. The "client", "terminal" and "terminal device" used herein may also be a communication terminal, an Internet terminal, a music / video playing terminal, for example, a PDA, a MID (Mobile Internet Device) and / or a mobile phone with a music / video playing function, or a smart TV, a set-top box and other devices.

[0064] The hardware referred to by the names such as "server", "client", and "service node" in this application is essentially an electronic device with the equivalent capabilities of a personal computer. It is a hardware device with the necessary components revealed by the von Neumann principle, such as a central processing unit (including an arithmetic unit and a controller), a memory, an input device, and an output device. The computer program is stored in its memory, and the central processing unit calls the program stored in the external memory into the internal memory for execution, executes the instructions in the program, and interacts with the input and output devices to complete specific functions.

[0065] It should be pointed out that the concept of "server" referred to in this application can also be extended to the case of server clusters. According to the network deployment principle understood by those skilled in the art, the servers should be logically divided. In physical space, these servers can be independent of each other but can be called through interfaces, or integrated into a physical computer or a set of computer clusters. Those skilled in the art should understand this flexibility, and should not use it to restrict the implementation of the network deployment method of this application.

[0066] Unless expressly specified, one or more technical features of the present application can be deployed on a server for implementation and accessed by a client through a remote call to obtain an online service interface provided by the server, or can be directly deployed and run on a client for access.

[0067] The neural network models referenced or may be referenced in this application, unless expressly specified, can be deployed on a remote server and remotely called on the client, or can be deployed and directly called on a client with sufficient device capabilities. In some embodiments, when it runs on the client, its corresponding intelligence can be obtained through transfer learning to reduce the requirements for the client's hardware operating resources and avoid excessive occupation of the client's hardware operating resources.

[0068] Unless explicitly specified, the various data involved in this application can be stored remotely on a server or on a local terminal device, as long as it is suitable for being called by the technical solution of this application.

[0069] Those skilled in the art should be aware that, although the various methods of the present application are described based on the same concept and thus present commonality to each other, unless otherwise specified, these methods can be independently executed. Similarly, for each embodiment disclosed in the present application, they are all proposed based on the same inventive concept, therefore, concepts with the same expression, and concepts that are appropriately changed for convenience despite different expressions, should be understood as equivalent.

[0070] Unless the mutually exclusive relationship between the embodiments to be disclosed in this application is explicitly stated, the relevant technical features involved in each embodiment can be cross-combined to flexibly construct a new embodiment, as long as such combination does not deviate from the creative spirit of this application and can meet the needs of the prior art or solve certain deficiencies in the prior art. Those skilled in the art should be aware of this flexibility.

[0071] A commodity information matching method of the present application can be programmed as a computer program product and deployed in a server for execution, so that the method can be executed by accessing an interface opened after the computer program product is executed and performing human-computer interaction with the computer program product through a graphical user interface.

[0072] See also Figure 1 The product information matching method of the present application comprises the following steps:

[0073] Step S1100, obtaining the title text of the product to be released submitted by the terminal device;

[0074] One of the application scenarios of the present application is that when a merchant user needs to put a new product on the shelf, it is necessary to enter the product information of the new product for listing and publishing. To this end, the product information of the product to be released is entered in the product information entry interface of the product to be released. The product information usually includes title text, detail text, product image, etc. The number of words corresponding to the detail text is relatively large, and direct writing is relatively slow, while inputting a title text with fewer words is relatively fast. Therefore, the present application allows the terminal device to submit the title text to the server to obtain the detail text of the historical product that matches the semantics of the title text as the detail information template of the product to be released.

[0075] In another application scenario of the present application, when a merchant user needs to publish a product as an advertisement object, he also needs to enter the product information of the corresponding advertisement product for publishing the product advertisement. To this end, the product information of the product to be published is entered in the product information entry interface of the product to be published. Similarly, it is usually required to enter title text, detail text, product image and other information. In order to facilitate the filling of detail text information, the terminal device is allowed to submit the title text to the server to obtain the detail text of historical products that are similar or related to the title text and have good advertising results, as the detail information template of the product to be published.

[0076] In the above scenarios, new products that need to be listed and products that need to be advertised can be accessed through mobile phones, computers and other terminal devices as pending products to be released. The product information can be entered and submitted to the server of the e-commerce platform after the product information is entered.

[0077] Step S1200: extracting deep semantic information of the title text to obtain a corresponding text feature vector;

[0078] In one embodiment, in order to improve the accuracy of semantic information, the title text may be formatted preprocessed. Specifically, the format of general product title text information is relatively neat, mostly in plain text, while the format of product detail text information is relatively messy. However, it is not ruled out that the product title may contain line breaks, links, emoticons, extra punctuation marks, extra blank characters, etc. In order to make the semantic extraction more accurate, these extra characters that do not represent the meaning of the product content can be formatted accordingly, similar to the detail text, and these useless characters can be removed so that the title text and detail text are unified in format and the character sets contained in each other are relatively close.

[0079] An exemplary partial preprocessing process includes: replacing line breaks and tabs with space symbols; replacing http connections with space symbols; replacing emoji expressions with space symbols; converting all English characters to lowercase; filtering comment format strings such as @xxx; replacing more than two blank strings with only one blank character; replacing more than two punctuation marks with only one punctuation mark, etc. The format preprocessing method is adopted as needed, and those skilled in the art can flexibly implement it according to actual business conditions.

[0080] Of course, if the format of the title text is relatively neat, there is no need to perform the above-mentioned formatting preprocessing.

[0081] After the server receives the title text, it can call the pre-trained text feature extraction model to extract deep semantic information from the title text, thereby obtaining the corresponding text feature vector. The text feature extraction model is pre-trained to a convergence state, and is used to perform deep semantic-based representation learning on the title text input therein. Those skilled in the art will appreciate that as long as a sufficient amount of samples are used to train the text feature extraction model so that the text feature extraction model is suitable for extracting deep semantic information from the title text, the text feature extraction model can be put into use in the technical solution of the present application. The text feature extraction model is generally a neural network model, preferably a model based on Transformer implementation, so as to be more suitable for processing sequence information such as text. Subsequent embodiments of the present application will further reveal the preferred model selection method, which will not be discussed here for the time being.

[0082] Step S1300, calculating the similarity between the text feature vector and the text feature vector of the detail text of the historical product in the product feature library, and sorting the detail text of the historical product according to the similarity;

[0083] After obtaining the text feature vector of the title text, similarity data can be calculated between it and the text feature vector of the detail text of each historical product in a preset product feature library to determine the similarity information between the title text submitted by the terminal device and the detail text of each historical product in the product feature library.

[0084] The commodity feature library includes a plurality of historical commodities, which are pre-published and / or pre-selected commodities. The commodity feature library stores the detailed text of each historical commodity and its corresponding text feature vector in association. Similarly, the text feature vector of the historical commodity is also pre-extracted by the text feature extraction model.

[0085] There are many ways to calculate the similarity between the text feature vector of the title text and the text feature vector of the historical product. For example:

[0086] In one embodiment, common algorithms such as cosine similarity algorithm, Euclidean distance algorithm, Pearson coefficient algorithm, Jaccard similarity algorithm, etc. can be used for calculation. Taking the cosine similarity algorithm as an example, the text feature vector of the title text and the detail text feature vector of a historical product are input into the cosine similarity algorithm function as method parameters, and then the preset cosine similarity calculation formula can be applied to calculate the product of the two text feature vectors divided by the product of the modules corresponding to the two vectors, and obtain the corresponding output result, which is the similarity between the two text feature vectors.

[0087] In another embodiment, the index of the text feature vector of the detail text of the historical product in the product feature library is constructed through the interface provided by the faiss framework, and further through the similarity calculation interface provided by the framework, the corresponding preset algorithm can be applied to obtain the similarity data, thereby quickly calculating the similarity between the title text of the product to be released and the detail text of the historical product.

[0088] It can be seen that the algorithm for achieving the similarity between the text feature vector of the title text and the text feature vector of the historical products in the product feature library can be flexibly set. When the similarity between a title text and the detail text of each historical product is calculated based on the text feature vector, a corresponding similarity sequence can be obtained. For easy search, the detail text of the historical product can be sorted in descending order according to the similarity.

[0089] Step S1400: Push the sorted detail texts to the terminal device for display as a list of candidate detail texts for the product to be released.

[0090] In order to provide a reference scheme for inputting the details text of the goods to be released to the terminal device, the details text of the sorted historical goods can be constructed into a list of candidate details texts and pushed to the information input interface of the goods to be released on the terminal device for display. The candidate details text list includes the details text of the candidate historical goods, and the candidate details text can be used as the details text template adopted by the goods to be released. As mentioned above, the details text in the product feature library can be formatted in advance and can be processed into a single line of data for easy display.

[0091] After receiving the list of alternative detail texts, the terminal device can display it on the input interface for the user to select. In this way, the user can obtain detail text templates of historical products with similar semantics. By clicking one of them, the user can use the corresponding detail text template as the detail text of the product to be released, and can further edit it on this basis until the user is satisfied.

[0092] In one embodiment, the terminal device can submit the title text to the server by touching a preset control, so as to trigger the server to execute steps S1100 to 1400 so that the terminal device can obtain the list of candidate detail texts. In an alternative embodiment, the terminal device may not need to trigger the execution of the server through the control, but trigger the server execution in the background after it captures the title text of the product to be released input in its interface, and similarly, the candidate detail texts can also be obtained.

[0093] This typical embodiment embodies the advantages of the present application by implementing semantic matching on the server side according to the text feature vector of the title text in the product information of the product to be released, calculating the similarity between the text feature vector and the text feature vector of the detail text in the product information of the historical products, determining the detail texts of multiple historical products semantically associated with the title text of the product to be released according to the similarity, and pushing them to the terminal device as an optional option for entering the detail text in the product information of the product to be released, thereby improving the intelligence level of the server's product entry service, and further improving the product information entry efficiency on the user side, providing convenience for quick release in application scenarios such as product advertising and online release.

[0094] Furthermore, the present application implements, on the terminal device side, obtaining a list of semantically related alternative detail texts from a server based on the title text entered by the user in the first input box corresponding to the product title of the entry interface for the product to be released, and using each detail text in the list as an optional option for the input content of the second input box of the entry interface for entering the detail text of the product to be released. As a result, the user only needs to enter the title text of the product title of the product to be released in the first input box to obtain optional detail text that is semantically similar to the title text in the second input box. The user can then quickly modify and edit the second input box of the product to be released based on the selection of a detail text, thereby realizing quick and convenient input of product details of the product to be released, improving the editing efficiency of the product information of the product to be released, and obtaining a good user experience.

[0095] In addition, in the business logic implemented in this application, semantic matching is performed between the title text of the product to be released and the detail text of the historical product, rather than semantic matching between title texts. Therefore, the alternative detail text can be directly hit through the title text. Since the two are semantically associated, the matching degree is higher and the hit is more accurate.

[0096] See also Figure 2 In a further embodiment, the step S1300, calculating the similarity between the text feature vector and the text feature vector corresponding to the detail text of the historical product in the product feature library, and sorting the detail text of the historical product according to the similarity, comprises the following steps:

[0097] Step S1310, calculating the similarity between the text feature vector of the title text of the to-be-released product and the text feature vector of the detail text of the historical products in the product feature library, and obtaining a historical product similarity sequence, wherein each text feature vector is extracted using the same pre-trained neural network model;

[0098] The pre-trained neural network model is a neural network model that can calculate a sentence to obtain its vectorized text feature vector. The open source framework Sentence Transformers can be used, which provides a large number of pre-trained Transformer models, such as BERT, RoBERTa, XLM-RoBERTaModel, MPNet, etc. In this embodiment, the XLMRobertaModel is recommended. The specific implementation method can be flexibly adapted by technicians in this field according to actual business scenarios.

[0099] The XLMRobertaModel model is used to extract the deep semantic features of the preprocessed title text of the product to be released to obtain the corresponding text feature vector. Similarly, the preprocessed historical product details text is also pre-extracted with the XLMRobertaModel model to obtain the corresponding text feature vector and store it in the product feature library.

[0100] Obtain a set of text feature vectors corresponding to the detail texts of historical products in the product feature library, and use the cosine similarity function to calculate the similarity between the text feature vectors corresponding to the title text of the product to be released, as a measure of the semantic similarity between the title text of the product to be released and the detail text of the historical product. The example formula is as follows:

[0101]

[0102] Among them: product_title_embedding is the text feature vector corresponding to the title text of the product to be released, and passage_detail_embedding is the text feature vector corresponding to the detail text of the historical product.

[0103] According to the formula, the cosine similarity is obtained by calculating the product of the text feature vectors corresponding to the title text of the product to be released and the detail text of the historical product, divided by the product of the moduli corresponding to their vectors. For the convenience of calculation, the value range of the cosine similarity is normalized to 0-1, with the limits being 0 and 1 respectively. The larger the cosine similarity value, the higher the semantic similarity between the two. Conversely, the smaller the value, the greater the difference between the two and the lower the semantic similarity.

[0104] At this point, the cosine similarity of each corresponding item in the detailed text of a group of historical products is calculated. The cosine similarity and the historical products are in a one-to-one mapping relationship, thereby forming a historical product similarity sequence.

[0105] Step S1320: Sort the historical products in the historical product similarity sequence whose similarity exceeds a preset threshold according to the similarity to obtain a first recommendation group;

[0106] For the products to be released, the cosine similarity of the representative values ​​of the title text and the detail text in the historical products at the semantic similarity level has been calculated, and the historical products whose cosine similarity in the historical product similarity sequence exceeds a preset similarity threshold, i.e., the historical products corresponding to the preset threshold, are screened and sorted in descending order of cosine similarity, thereby constructing a first recommendation group, and obtaining the detail texts of multiple historical products with a high degree of semantic similarity to the title text of the product to be released as the first recommendation group.

[0107] The preset similarity threshold is 0.8. This embodiment recommends that the similarity threshold be selected and set within the range of 0.3-0.8. The specific value of the preset threshold can be flexibly adjusted by technicians in this field according to the actual business situation.

[0108] Step S1340: construct a detailed text recommendation list for pushing to the terminal device for display, and the detailed text recommendation list includes the first recommendation group.

[0109] In one embodiment, the detail text of each historical product in the first recommendation group is obtained to construct a detail text recommendation list, that is, an alternative detail text list, and each detail text is pushed to the terminal device as an option content of the detail text recommendation list. As a result, the user of the terminal device can roughly understand the corresponding detail text content by browsing the detail text recommendation list options, and can make a selection based on this to obtain the detail text content of the corresponding historical product. The detail text recommendation list is associated with a recommendation copy control, and when the user clicks the recommendation plan control, it triggers the server to execute the corresponding steps of the method and push the detail text recommendation list to the terminal device's product information entry interface for display.

[0110] In this embodiment, a recommendation list is constructed by using the detail texts of historical products that are semantically similar to the title text of the product to be released, and an intelligent product information entry service is provided for users to input the detail texts of the product to be released. Existing resources are used more rationally, so that users can obtain semantically similar detail text templates as references through simple operations, which makes it easier for users to write large-scale detail texts more quickly. Users can also gain inspiration from the various templates and create better works based on the selected template, thereby improving user experience and product information entry efficiency.

[0111] See also Figure 3 In an extended embodiment, after the step S1320, according to the similarity, sorting the historical commodities in the historical commodity similarity sequence whose similarity exceeds a preset threshold value to obtain the first recommendation group, the following steps are also included:

[0112] Step S1330: Sort the historical products in the historical product similarity sequence whose similarity is lower than a preset threshold value according to the advertising data of the historical products to obtain a second recommendation group, and the second recommendation group is connected after the first recommendation group in the detail text recommendation list.

[0113] Filter out the historical products with similarity lower than a preset threshold in the historical product similarity sequence, obtain the latest advertisement data of the historical products in the advertisement product database, quantify the advertisement data to obtain corresponding scores, and then rearrange the order of the historical products from high to low according to the scores, and further construct a second recommendation group with the detail texts of the sorted historical products. Put the first recommendation group first and the second recommendation group second, and splice them together successively to construct and expand the detail text recommendation list, and then push the detail text recommendation list to the product release interface.

[0114] In this embodiment, the user first chooses to use the detail text in the first recommendation group that is highly similar in semantics to the title text he has set as a template to quickly complete the creation of a large amount of detail text. Secondly, the user can also choose the detail text in the second recommendation group that is less similar in semantics to the title he has set but has better advertising effect as a template. The latter has the advantages of clear description and popularity among the public. Creation on this basis can achieve better advertising effects.

[0115] See also Figure 4 In a specific embodiment, the step S1330, according to the advertisement data of the historical products, sorting the historical products in the historical product similarity sequence whose similarity is lower than a preset threshold to obtain the second recommendation group, comprises the following steps:

[0116] Step S1331, extracting historical products whose similarity is lower than a preset threshold from the historical product similarity sequence to obtain a temporary calculation list;

[0117] The historical product similarity sequence is constructed by calculating the cosine similarity based on the text feature vector of the title text of the product to be released and the text feature vector of the detail text of the historical product and sorting them from high to low. Therefore, according to the similarity sorting of the historical product similarity sequence and the corresponding cosine similarity, the historical products with cosine similarity lower than the preset threshold are screened out to obtain a temporary calculation list.

[0118] Step S1332, calling multiple pieces of advertising data of each historical product in the temporary calculation list, wherein the advertising data includes any one or multiple pieces of the corresponding historical product advertising time, advertising time, and advertising effectiveness data;

[0119] Call multiple predetermined advertising data for the historical products in the temporary calculation list. The advertising data includes relevant data for analyzing the advertising time and corresponding effectiveness of the historical products, which can be called from the advertising product library storing these data. For example, the advertising data includes: advertising duration, which refers to the number of days from the last advertising date of the historical product to the current date; advertising duration, which refers to the number of days from the last advertising date of the historical product; advertising effectiveness data, including the comprehensive quantitative data calculated by adding the click rate, purchase rate, and ROAS (return on advertising expenditure) obtained by the most recent advertising of the historical product. This is only an example, and those skilled in the art can also flexibly set these advertising data according to actual needs.

[0120] In order to better analyze the latest advertising data of the historical products and comprehensively evaluate the advertising effectiveness of the historical products, the advertising data can be normalized to obtain advertising data under the same scale specification. In one embodiment, the advertising duration, advertising duration, and advertising effectiveness data of the historical products in the temporary calculation list are traversed, and the maximum and minimum values ​​of the various advertising data are obtained respectively. Then, the current advertising data of the historical products are calculated by subtracting the corresponding maximum value and dividing by the difference between the corresponding maximum value and the minimum value to obtain the normalized result, that is, the current advertising data is linearly transformed using Min-Max standardization, and the normalized result is mapped to between 0 and 1. The example formula is as follows:

[0121]

[0122] Wherein: y is the normalized result of each advertisement data; x is the value corresponding to each advertisement data; min{x} is the minimum value among the values ​​corresponding to each advertisement data; and max{x} is the maximum value among the values ​​corresponding to each advertisement data.

[0123] The normalization method is not limited to the min-max normalization used in the current embodiment, and other normalization methods such as truncation, binning, logarithmic transformation, Box-Cox, Z-Score normalization, etc. may also be used. Those skilled in the art can flexibly select and use them according to the actual score distribution.

[0124] Step S1333: for each historical product in the temporary calculation list, weighted sum of multiple pieces of advertising data is performed to obtain a corresponding temporary score;

[0125] Calculate the normalized advertising duration, advertising duration, and advertising effectiveness data of each historical product in the temporary calculation list, multiply them by the corresponding weights, and add them up to obtain the corresponding temporary score. The exemplary formula is as follows:

[0126] score = a*x 1 +b*x 2 +c*x 3

[0127] Among them: score is the temporary score, x 1 is the current length of the advertisement, x 2 is the data value of the advertising duration, x 3 is the advertising effectiveness data value, and a, b, c are the weights corresponding to each advertising data value.

[0128] In one embodiment, the weight corresponding to the advertisement duration data value is 0.3, the weight corresponding to the advertisement delivery duration data value is 0.3, and the weight corresponding to the advertisement delivery effectiveness data value is 0.4. The specific weight setting can be flexibly adapted by those skilled in the art according to the actual situation.

[0129] Step S1334: Sort the historical products in the temporary calculation list according to the temporary scores to obtain the second recommendation group.

[0130] The historical commodities in the temporary calculation list are sorted accordingly according to the temporary scores of the historical commodities, and a second recommendation group is further constructed according to the detail texts and title texts of the sorted historical commodities.

[0131] The second recommendation group is located immediately after the first recommendation group, and together with the first recommendation group, constitutes the detailed text recommendation list.

[0132] In this embodiment, the latest advertising data of historical products is fully analyzed to achieve deep data mining, and the advertising data is further quantified to calculate the corresponding score and then construct the second recommendation group based on it, and the second recommendation group is used to construct the detailed text recommendation list. It can be seen that this embodiment rationally uses existing resources, scientifically and accurately quantifies the latest advertising results of existing resources into corresponding scores, and makes the detailed text provided in the detailed text recommendation list more valuable for reference, so that users can quickly and easily obtain the detailed text of historical products with good advertising as a reference, which greatly improves the user experience and can expect the products to be released to obtain good advertising effects.

[0133] A commodity information publishing method of the present application can be programmed as a computer program product and deployed in a client for execution, so that the method can be executed by accessing an interface opened after the computer program product is executed and performing human-computer interaction with the computer program product through a graphical user interface.

[0134] See also Figure 5 The product information publishing method of this application includes the following steps:

[0135] Step S2100: displaying a product information input interface of the product to be released, the input interface including a first input box corresponding to the title text in the product information and a second input box corresponding to the detail text;

[0136] See also Figure 8 On the terminal device as the client, the product information entry interface of the product to be released includes a first input box 100 corresponding to the title text in the product information and a second input box 110 corresponding to the detail text, as well as a recommended copy control 120 for recommending a list of alternative detail texts. The user enters the title text of the product to be released in the first input box 100 on the interface.

[0137] Step S2200: In response to a preset trigger event acting on the second input box, submit the title text in the first input box to the server to obtain a list of candidate detail texts semantically matching the title text, the list including detail texts corresponding to a plurality of historical products;

[0138] See also Figure 8 , when the user clicks the recommended text control 120 on the product information entry interface of the product to be released, the preset trigger event is triggered, thereby submitting the title text entered in the first input box to the server. The server executes the product information matching method of the present application, obtains the corresponding alternative detail text list based on the title text in the first input box 100, and pushes it to the terminal device for display. As mentioned above, the detail text in the alternative detail text list is the historical product detail text that semantically matches the title text in the first input box. Step S2300, output the alternative detail text list to the current entry interface to display the detail text from multiple historical products, so that each detail text is added to the second input box in response to the touch operation;

[0139] See also Fig. 9, the terminal device outputs the candidate detail text list 130 to the detail text input box of the product to be released in the interface to display the title text options of multiple historical products. After the user browses the options and knows the content of the detail text that matches the semantics of the title text of the product to be released, the user selects the appropriate detail text. In response to the user's touch operation on any option in the candidate detail text, the terminal device adds the detail text selected by the user to the second input box 110 as the detail text template of the product to be released for further editing by the user.

[0140] Step S2400: In response to the information submission event acting on the input interface, the product information displayed in the first input box and the second input box is submitted to the server to execute the release of the product to be released.

[0141] See also Fig. 9 , after completing the title text of the first input box 100 and the detail text of the second input box 110, as well as other product information, the user can click the submit control 140 to publish on the product information entry interface. In response to the submit event acting on the submit control 140, the terminal device submits all product information of the product to be published to the server, and the server publishes the product to be published. Accordingly, it can adapt to different application scenarios and realize the listing of new products or the publication of product advertisements.

[0142] In this embodiment, on the interface for publishing product information entry, recommended copy of alternative detail text that matches the semantics of the title text is provided. A high-quality detail text template can be obtained by simply clicking to select, thereby quickly completing the entry of product information, improving user experience and increasing user stickiness.

[0143] See also Figure 6 In a preferred embodiment, the preset trigger event acting on the second input box is a text input event generated by inputting text in the first input box.

[0144] See also Fig.10 The terminal device can monitor the input event of the user entering the title text in the first input box 200 on the information entry interface of the product to be released, and regard it as the preset trigger event. Therefore, the corresponding trigger submits the title text to the server, thereby obtaining the alternative detail text list pushed by the server according to the title text, and the alternative detail text list 220 is displayed on the interface. The displayed alternative text list options are the abbreviation texts of the title text entered in the first input box as the keyword in the corresponding detail text. The detail text option selected by the user is added to the second input box 210 as the detail text template of the product to be released.

[0145] In this embodiment, an intelligent template service is provided for the writing of long-length detailed text, which simplifies a certain amount of input, allowing users to quickly complete the most complicated steps in entering product information, thereby improving the efficiency of entering product information.

[0146] See also Figure 7 In an extended embodiment, after the step S2300, outputting the candidate detail text list to the current input interface to display detail texts from multiple historical commodities, so that each detail text is added to the second input box in response to a touch operation, the following steps are included:

[0147] Step S2301: In response to a selection event acting on any detail text in the candidate detail text list, the selected detail text is connected to the input content in the second input box and displayed.

[0148] Click to select an option in the alternative detail text list, and in response to the selection event, add the selected detail text to the content already entered in the second input box and display it. If multiple options in the alternative detail text list are selected, the selected detail texts will be added to the content already in the second input box in the order of selection time. In this way, it can avoid affecting the detail text that the user has edited on the interface.

[0149] In this embodiment, multiple detail text templates can be used for reference in further creation of the product detail text to be released, making the writing of the detail text simpler and having better results.

[0150] See also Fig.11 , a product information matching device provided to meet one of the purposes of the present application, includes: a title acquisition module 1100, a semantic extraction module 1200, a similarity matching module 1300, and a push display module 1400, wherein the title acquisition module is used to obtain the title text of the product to be released submitted by the terminal device; the semantic extraction module is used to extract the deep semantic information of the title text to obtain the corresponding text feature vector; the similarity matching module is used to calculate the similarity between the text feature vector and the text feature vector of the detail text of the historical product in the product feature library, and sort the detail text of the historical product according to the similarity; the push display module is used to push the sorted detail text to the terminal device for display as a list of alternative detail texts for the product to be released.

[0151] In a further embodiment, the similarity calculation module includes: a similarity calculation submodule, which is used to calculate the similarity between the text feature vector of the title text of the product to be released and the text feature vector of the detail text of the historical product in the product feature library, and obtain a historical product similarity sequence, and each of the text feature vectors is extracted using the same pre-trained neural network model; a similarity selection submodule, which is used to sort the historical products in the historical product similarity sequence whose similarity exceeds a preset threshold according to the similarity, and obtain a first recommendation group; a list push submodule, which is used to construct a detail text recommendation list for pushing to the terminal device for display, and the detail text recommendation list includes the first recommendation group.

[0152] In an extended embodiment, the similarity calculation module also includes: a historical preference sub-module, which is used to sort the historical products in the historical product similarity sequence whose similarity is lower than a preset threshold according to the advertising data of the historical products, and obtain a second recommendation group, and the second recommendation group is connected after the first recommendation group in the detail text recommendation list.

[0153] In a specific embodiment, the historical optimization submodule includes: a low-value extraction unit, which is used to extract historical products with a similarity lower than a preset threshold from the historical product similarity sequence to obtain a temporary calculation list; a data calling unit, which is used to call multiple advertising data of each historical product in the temporary calculation list, and the advertising data includes any one or multiple of the corresponding historical product advertising duration, advertising duration, and advertising effectiveness data; a comprehensive quantification unit, which is used to weightedly sum multiple advertising data for each historical product in the temporary calculation list to obtain a corresponding temporary score; a sorting processing unit, which is used to sort the historical products in the temporary calculation list according to the temporary score to obtain the second recommendation group.

[0154] See also Fig.12, a product information publishing device provided to meet one of the purposes of the present application, includes: an interface display module 2100, an alternative acquisition module 2200, a text display module 2300, and a submission and publishing module 2400, wherein the interface display module is used to display a product information entry interface of a product to be released, and the entry interface includes a first input box corresponding to a title text in the product information and a second input box corresponding to a detail text; the alternative acquisition module is used to respond to a preset trigger event acting on the second input box, submit the title text in the first input box to a server to obtain a list of alternative detail texts semantically matching the title text, and the list includes detail texts corresponding to multiple historical products; the text display module is used to output the list of alternative detail texts to the current entry interface to display detail texts from multiple historical products, so that each detail text is added to the second input box in response to a touch operation; the submission and publishing module is used to respond to an information submission event acting on the entry interface, and submit the product information displayed in the first input box and the second input box to the server to execute the release of the product to be released.

[0155] In a preferred embodiment, the preset trigger event acting on the second input box is a text input event generated by inputting text in the first input box.

[0156] In an extended embodiment, the product information publishing device also includes: a selection insertion module, which is used to respond to a selection event acting on any detail text in the alternative detail text list, and connect the selected detail text to the input content in the second input box and then display it.

[0157] In order to solve the above technical problems, the present application also provides a computer device. Fig.13 As shown, a schematic diagram of the internal structure of a computer device. The computer device includes a processor, a computer-readable storage medium, a memory, and a network interface connected via a system bus. Among them, the computer-readable storage medium of the computer device stores an operating system, a database, and computer-readable instructions. The database may store a control information sequence. When the computer-readable instructions are executed by the processor, the processor may implement a commodity information matching method. The processor of the computer device is used to provide computing and control capabilities to support the operation of the entire computer device. The memory of the computer device may store computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor may execute the commodity information matching method of the present application. The network interface of the computer device is used to connect and communicate with a terminal. Those skilled in the art will understand that Fig.13The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0158] In this embodiment, the processor is used to execute Fig.11 , 12 The memory stores the program code and various data required to execute the above modules or submodules. The network interface is used to transmit data between user terminals or servers. The memory in this embodiment stores the program code and data required to execute all modules / submodules in the commodity information matching device of this application, and the server can call the program code and data of the server to execute the functions of all submodules.

[0159] The present application also provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the commodity information matching method of any embodiment of the present application.

[0160] The present application also provides a computer program product, including a computer program / instruction, which implements the steps of the method described in any embodiment of the present application when executed by one or more processors.

[0161] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments of the present application can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a computer-readable storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0162] To sum up, this application facilitates the entry of product information of products to be released, can intelligently recommend detailed text of product information based on the title text of the product information that has been entered, can serve a variety of application scenarios including the e-commerce field, and has wide versatility.

[0163] Those skilled in the art will appreciate that the various operations, methods, steps, measures, and schemes in the processes discussed in this application may be alternated, altered, combined, or deleted. Further, other steps, measures, and schemes in the various operations, methods, and processes discussed in this application may also be alternated, altered, rearranged, decomposed, combined, or deleted. Further, the steps, measures, and schemes in the prior art that are similar to those disclosed in this application may also be alternated, altered, rearranged, decomposed, combined, or deleted.

[0164] The above description is only a partial implementation method of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A commodity information matching method, It is characterized in that The steps include: Obtain the title text of the product to be released submitted by the terminal device; Extracting deep semantic information of the title text to obtain a corresponding text feature vector; Calculating the similarity between the text feature vector and the text feature vector of the detail text of the historical product in the product feature library, and sorting the detail text of the historical product according to the similarity; Pushing the sorted detail texts to the terminal device for display as a list of candidate detail texts for the products to be released; The calculating the similarity between the text feature vector and the text feature vector corresponding to the detail text of the historical commodity in the commodity feature library, and sorting the detail text of the historical commodity according to the similarity, comprises the following steps: Calculating the similarity between the text feature vector of the title text of the product to be released and the text feature vector of the detail text of the historical products in the product feature library to obtain a historical product similarity sequence, wherein each of the text feature vectors is extracted using the same pre-trained neural network model; According to the similarity, the historical products in the historical product similarity sequence whose similarity exceeds a preset threshold are sorted to obtain a first recommendation group; According to the advertisement data of the historical products, the historical products whose similarity is lower than a preset threshold in the historical product similarity sequence are sorted to obtain a second recommendation group, and the second recommendation group is connected to the first recommendation group in the detail text recommendation list; The detailed text recommendation list is constructed for being pushed to the terminal device for display, and the detailed text recommendation list includes the first recommendation group and the second recommendation group.

2. The commodity information matching method according to claim 1, It is characterized in that According to the advertisement data of historical products, historical products with similarity lower than a preset threshold in the historical product similarity sequence are sorted to obtain a second recommendation group, including the following steps: Extract historical products whose similarity is lower than a preset threshold from the historical product similarity sequence to obtain a temporary calculation list; Calling multiple pieces of advertising data of each historical product in the temporary calculation list, wherein the advertising data includes any one or multiple pieces of the corresponding historical product advertising time, advertising time, and advertising effectiveness data; For each historical product in the temporary calculation list, weighted summing up multiple pieces of advertising data to obtain a corresponding temporary score; The historical commodities in the temporary calculation list are sorted according to the temporary scores to obtain the second recommendation group.

3. A method for publishing product information, It is characterized in that The steps include: Displaying a product information input interface for the product to be released, the input interface including a first input box corresponding to the title text in the product information and a second input box corresponding to the detail text; In response to a preset trigger event acting on the second input box, submitting the title text in the first input box to the server to obtain a list of candidate detail texts semantically matching the title text, the list including detail texts corresponding to a plurality of historical commodities, wherein the preset trigger event acting on the second input box is a text input event generated by inputting text in the first input box; Outputting the candidate detail text list to the current input interface to display detail texts from a plurality of historical commodities, so that each detail text is added to the second input box in response to the touch operation; In response to the information submission event acting on the input interface, the product information displayed in the first input box and the second input box is submitted to the server to execute the release of the product to be released.

4. The method for publishing product information according to claim 3, It is characterized in that After the step of outputting the candidate detail text list to the current input interface to display detail texts from a plurality of historical commodities so that each detail text is added to the second input box in response to the touch operation, the following steps are included: In response to a selection event acting on any detail text in the candidate detail text list, the selected detail text is connected to the input content in the second input box and displayed.

5. A computer device comprising a central processing unit and a memory, It is characterized in that The central processing unit is used to call and run the computer program stored in the memory to execute the steps of the method according to any one of claims 1 to 4.

6. A computer-readable storage medium, It is characterized in that It stores a computer program implemented according to the method described in any one of claims 1 to 4 in the form of computer-readable instructions, and when the computer program is called and executed by a computer, the steps included in the corresponding method are executed.

7. A computer program product comprising a computer program / instructions, It is characterized in that When the computer program / instructions are executed by a processor, the steps of the method described in any one of claims 1 to 4 are implemented.

Images