Commodity library information generation method and device, electronic equipment and storage medium

Abstract

The invention discloses a commodity library information generation method and device, electronic equipment and a storage medium. The method comprises the following steps: determining commodity basic attribute information of a to-be-processed commodity; according to the commodity basic attribute information and the copywriting attribute information, determining a prompt text of a large language model, and inputting the prompt text into the large language model to obtain commodity copywriting information; and taking the commodity basic attribute information and the commodity copywriting information as commodity library information matched with the to-be-processed commodity. According to the technical scheme, the commodity copywriting information is obtained through the commodity basic attribute information, the copywriting attribute information and the large language model, the commodity library information is determined according to the commodity basic attribute information and the commodity copywriting information, the commodity library information corresponding to the commodity is automatically generated, the commodity library maintenance cost and the commodity input difficulty are reduced, and the commodity input efficiency is improved. And the commodity input efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method, device, electronic device and storage medium for generating commodity library information. Background Art

[0002] When a store uses a cash register system to sell goods, it first needs to enter the product information into the cash register system. For standard products, you can query the barcode library to obtain unified product information. However, for non-standard products, there is currently no similar barcode library to meet the needs of fast product entry. In addition, the description and display effects of non-standard products require manual creation to produce relevant information, which has high requirements for personnel and cannot guarantee continuous high-quality output. When the products are frequently replaced (such as clothing), the store will spend a lot of time on creating the product library.

[0003] In the existing methods, non-standard product databases are usually entered one by one or through file import. However, there are many items to be entered one by one, such as specifications, sizes, and colors, and the maintenance of the product database is large. This method will increase the processing time and efficiency of non-standard product database entry. Among them, the processing of files needs to be carried out according to the requirements of the software. Format errors or mislisting of contents will cause database construction failure, which makes processing difficult. In addition, the description of product characteristics varies from person to person, and it is impossible to guarantee high-quality output.

[0004] Therefore, a solution is needed to solve the problem of insufficient data entry in the existing non-standard product library. Summary of the invention

[0005] The present invention provides a method, device, electronic device and storage medium for generating commodity library information, so as to realize automatic generation of commodity library information corresponding to commodities.

[0006] In a first aspect, an embodiment of the present invention provides a method for generating commodity library information, the method comprising:

[0007] Determine the basic attribute information of the goods to be processed;

[0008] Determine a prompt text of a large language model according to the basic attribute information of the product and the attribute information of the copy, and input the prompt text into the large language model to obtain the product copy information;

[0009] The basic attribute information of the commodity and the commodity text information are used as commodity library information matching the commodity to be processed.

[0010] In a second aspect, an embodiment of the present invention further provides a device for generating commodity library information, the device comprising:

[0011] A commodity attribute information determination module, used to determine the commodity basic attribute information of the commodity to be processed;

[0012] A commodity copy information determination module, used to determine the prompt text of the large language model according to the commodity basic attribute information and the copy attribute information, and input the prompt text into the large language model to obtain the commodity copy information;

[0013] The commodity library information determination module is used to use the commodity basic attribute information and the commodity copy information as commodity library information matching the commodity to be processed.

[0014] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a method for generating product library information as described in any one of the embodiments of the present invention is implemented.

[0015] In a fourth aspect, an embodiment of the present invention further provides a storage medium storing computer executable instructions, wherein the computer executable instructions, when executed by a computer processor, are used to execute a method for generating product library information as described in any one of the embodiments of the present invention.

[0016] The technical solution of the embodiment of the present invention obtains product copy information through a large language model, and determines product library information based on basic product attribute information and product copy information, thereby automatically generating product library information corresponding to the product, reducing product library maintenance costs and product entry difficulty, and improving product entry efficiency.

[0017] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 is a flow chart of a method for generating commodity library information provided by Embodiment 1 of the present invention;

[0020] Figure 2 is a flow chart of a method for generating commodity library information provided by Embodiment 2 of the present invention;

[0021] Figure 3This is a schematic diagram of a process for constructing product library information provided by Embodiment 2 of the present invention;

[0022] Figure 4 This is a schematic diagram of a module of commodity library information provided by Embodiment 2 of the present invention;

[0023] Figure 5 It is a structural schematic diagram of a device for generating commodity library information provided by Embodiment 3 of the present invention;

[0024] Figure 6 It is a structural schematic diagram of an electronic device provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 should fall within the scope of protection of the present invention.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. In the embodiments of the present application, some existing solutions in the industry such as certain software, components, models, etc. may be mentioned, and they should be considered as exemplary, and their purpose is only to illustrate the feasibility of the implementation of the technical solution of the present application, but it does not mean that the applicant has or must use the solution.

[0027] The acquisition, transmission, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0028] Embodiment 1

[0029] Figure 1A flowchart of a method for generating product library information is provided for the first embodiment of the present invention. This embodiment is applicable to the generation of product library information. The method can be executed by a device for generating product library information. The device for generating product library information can be implemented in the form of hardware and / or software. The device for generating product library information can be configured in any electronic device with network communication and computing capabilities. Figure 1 As shown, the method includes:

[0030] S110: Determine basic attribute information of the commodity to be processed.

[0031] In this embodiment, the basic attribute information of the commodity includes but is not limited to the appearance attribute information, material attribute information or label attribute information of the commodity.

[0032] The basic attribute information of the commodity may be determined by obtaining the commodity input information in response to the user interaction interface, or by automatically generating commodity information corresponding to the commodity to be processed through a tool.

[0033] As an optional but non-limiting implementation method, determining the basic attribute information of the commodity to be processed includes steps A1-A3:

[0034] Step A1: Determine the product image to be processed.

[0035] Step A2: performing image recognition on the commodity image to be processed to obtain at least one feature data.

[0036] Step A3: Determine the basic attribute information of the commodity to be processed according to the mapping relationship between the feature data and the preset feature data and commodity attributes.

[0037] The commodity image to be processed in this embodiment means an image containing the commodity to be processed, wherein the commodity image to be processed may be determined by obtaining image information recorded in real time by a camera device, or obtaining an image file submitted through a user interactive program interface.

[0038] Furthermore, image recognition is performed on the image of the commodity to be processed to obtain feature data of the commodity to be processed. Taking the commodity to be processed as clothing as an example, the feature data of the commodity to be processed includes clothing color, clothing shape and clothing texture.

[0039] It should be noted that commonly used image recognition methods for feature extraction include image recognition methods based on template matching, image recognition methods based on feature matching, and image recognition methods based on deep learning.

[0040] In this embodiment, the commodity basic attribute information of the commodity to be processed is determined according to the mapping relationship between the preset feature data and the commodity attributes.

[0041] Taking clothing as an example, the mapping relationship between the pre-set feature data and the product attributes includes the mapping relationship between RGB (red, green, blue) values ​​and clothing colors, the mapping relationship between local features (local edges, local shapes, textures) and local clothing attributes (notched lapels, herringbone patterns), and the mapping relationship between global features (overall edges, overall shapes) and clothing style information (suits, ties, shirts).

[0042] In practical applications, the residual network or convolutional neural network in the deep learning model can be used to perform image recognition on the product images to be processed. When learning image recognition, the image classification network can be used as a reference to first extract the pixel space of the image and then convert it into a higher-level concept space.

[0043] Among them, image recognition can be first performed on the clothing image to obtain RGB values, and color analysis can be performed on the image according to the preset mapping relationship between RGB values ​​and clothing colors, and the RGB values ​​can be converted into color attribute information of the product to be processed to obtain the product (clothing) color.

[0044] Specifically, the local features of texture can be converted into the herringbone pattern attribute information of the product to be processed according to the mapping relationship between local features (local edge, local shape, texture) and local attributes of clothing (notched lapel, herringbone pattern). Or, the global features of overall shape can be converted into the style attribute information of the product to be processed according to the mapping relationship between global features (overall edge, overall shape) and clothing style information (suit, tie, shirt).

[0045] It should be noted that the similarity between the product image to be processed and the pre-stored product image with an existing label may also be calculated to determine the basic product attribute information of the product to be processed.

[0046] After calculating the similarity between the image of the product to be processed and each product image with a label, if the similarity with a product image with a label is greater than or equal to the threshold, the product to be processed and the labeled product are considered to belong to the same category or have similar attributes. For example, if the similarity between the image of the product to be processed and a product image labeled as "sports shoes" exceeds the threshold, it can be preliminarily determined that the product to be processed is also a sports shoe, and then its basic attribute information, such as color, style, etc., can be obtained.

[0047] Furthermore, the basic attribute information of the commodity to be processed may be modified or supplemented in response to the input information submitted by the user interaction interface.

[0048] As an optional but non-limiting implementation method, determining the basic attribute information of the commodity to be processed further includes steps B1-B3:

[0049] Step B1: Determine the product image to be processed.

[0050] Step B2: Perform text recognition on the commodity image to be processed to obtain at least one item of text information.

[0051] Step B3: convert the text information into structured data as basic attribute information of the commodity to be processed.

[0052] The image to be processed in this embodiment represents an image containing a text label of the commodity to be processed, wherein the commodity image to be processed can be determined by obtaining image information recorded in real time by a camera device, or obtaining an image file submitted through a user interactive program interface.

[0053] It should be noted that the commodity images to be processed in different processing steps in this embodiment may be the same image or different image files, and are not specifically limited.

[0054] Furthermore, the commodity image to be processed can be recognized by using text recognition technology to obtain text information in the commodity image to be processed. Commonly used text recognition technologies include OCR (Optical Character Recognition), text recognition technology based on template matching, and text recognition technology based on feature extraction.

[0055] Specifically, OCR optical character recognition technology can be used to extract text features from text labels in the image of the product to be processed to determine the text information corresponding to the product to be processed. The text information is semantically recognized to determine the field corresponding to the text information, and the text information is converted into corresponding structured data. Taking the text information of the product to be processed as pure wool worsted fabric, men's clothing, spring and autumn as an example, the corresponding structured data is material (pure wool worsted fabric), category (men's clothing), season (spring and autumn), and the structured data is used as the basic attribute information of the product to be processed.

[0056] Furthermore, the basic attribute information of the commodity to be processed may be modified or supplemented in response to the input information submitted by the user interaction interface.

[0057] S120: Determine a prompt text of a large language model according to the basic attribute information of the product and the text attribute information, and input the prompt text into the large language model to obtain product text information.

[0058] In this embodiment, the text attribute information indicates the attribute information of the product description text, including the product text style, product text tone, and product keywords. The prompt text (also called prompt word, hereinafter referred to as prompt text) indicates the text information determined by the basic attribute information of the product and the text attribute information. The product text information indicates the information obtained by processing the prompt text through the large language model.

[0059] It should be noted that the structure of the prompt text may include instructions, context information, input data, output instructions, etc. In this embodiment, the instruction may be determined to be outputting the product copy, and the basic attribute information of the product and the copy attribute information may be directly used as context information to form the prompt text and input it into the large language model to obtain the product copy information. The instruction may also be determined to be outputting the product copy, and the basic attribute information of the product may be used as context information, and the copy attribute information may be used as an output instruction to form the prompt text and input it into the large language model. In this embodiment, the specific method of forming the prompt text according to the basic attribute information of the product and the copy attribute information is not specifically limited.

[0060] It needs to be explained that by determining the prompt text and inputting it into the large language model through the basic attribute information of the product and the copy attribute information, the product copy information can be obtained, which can realize the automatic generation of the product copy, improve the efficiency of generating the product copy information, and ensure the quality of the product copy.

[0061] S130: Use the basic attribute information of the commodity and the commodity text information as commodity library information matching the commodity to be processed.

[0062] In this embodiment, the commodity library information represents the warehouse information of the commodity to be processed, including the basic attribute information of the commodity and the commodity copy information.

[0063] In actual applications, the product library information may include the product name, number, brand, color, style, category, size, product introduction or detailed description information, etc.

[0064] As an optional but non-limiting implementation method, after determining the basic attribute information of the commodity to be processed, it also includes:

[0065] Inputting the basic attribute information of the commodity into the large language model to obtain the attribute information of the commodity after processing;

[0066] The basic attribute information of the commodity and the commodity text information are used as commodity library information matching the commodity to be processed, including:

[0067] The basic attribute information of the commodity, the attribute information of the commodity after processing and the commodity copy information are used as commodity library information matching the commodity to be processed.

[0068] In this embodiment, the processed attribute information of the commodity refers to the attribute information obtained by processing the basic attribute information of the commodity through the large language model. Taking clothing as an example, the basic attribute information of the commodity can be the color, style, material and other information directly obtained through image processing or text recognition; the processed attribute information of the commodity can be the name of the commodity, applicable season and other information obtained by further processing the basic attribute information of the commodity by the large language model.

[0069] The basic attribute information of the commodity is input into the large language model, and the attribute information of the commodity is directly generated according to the processing rules of the large language model to obtain the attribute information of the processed commodity. Exemplarily, the prompt text for the generated commodity name can be obtained according to the basic attribute information of the commodity, such as color, style, pattern, etc., and the instruction to instruct the large language model to generate the commodity name according to the basic attribute information of the commodity. The prompt text for the generated commodity name is input into the large language model to obtain the commodity name output by the large language model.

[0070] Furthermore, the basic attribute information of the commodity, the attribute information of the commodity after processing, and the commodity copy information are stored as commodity library information matching the commodity to be processed.

[0071] The technical solution of the embodiment of the present invention obtains product copy information through a large language model through product basic attribute information and copy attribute information, and determines product library information based on the product basic attribute information and product copy information, thereby automatically generating product library information corresponding to the product, reducing the product library maintenance cost and the difficulty of product entry, and improving product entry efficiency.

[0072] Embodiment 2

[0073] Figure 2 This is a flow chart of a method for generating product library information provided by Embodiment 2 of the present invention. This embodiment of the present invention is further specific based on the above embodiment. Figure 2 As shown, the method includes:

[0074] S210: Determine basic attribute information of the commodity to be processed.

[0075] It should be noted that this step has been described in the above embodiment and will not be repeated in this embodiment.

[0076] S220: Determine input information of the user interaction interface.

[0077] The input information includes product copywriting style, product key attributes, etc. input through the user interaction interface.

[0078] In this embodiment, when the input information of the user interaction interface is determined, the input information may also be verified to ensure that the input information is associated with the commodity and to ensure the validity of the input information.

[0079] S230: Process the input information by prompting the learner to obtain the text attribute information.

[0080] In this embodiment, the prompt learner is based on the big language model. The big language model has learned rich language knowledge and semantic representation on a large amount of text data. The prompt learner is a model or algorithm that can automatically learn and generate prompt texts. The prompt texts learned and generated by the prompt learner are usually some guiding texts, which can guide the big language model to make predictions in the desired direction, better understand the task and generate more accurate answers.

[0081] In this embodiment, in order to ensure the semantic consistency between the large language model and the decoder output, the prompt learner can process the input information and convert the input information into text attribute information that is suitable for constituting the large language model prompt text together with the product attribute information.

[0082] In this embodiment, the prompt text learning link is added mainly to ensure the generation effect of the large language model. It is an intervention link that can control the deviation of the generated product copy information and enrich the content of the product copy information, ensuring that the effect of the product copy information generated by the large voice model can be more in line with user expectations and satisfaction.

[0083] S240: Determine a prompt text of a large language model according to the basic attribute information of the product and the attribute information of the copy.

[0084] It should be noted that this step has been described in the above embodiment and will not be repeated in this embodiment.

[0085] S250: Input the prompt text into a large language model to obtain product copy information.

[0086] It should be noted that this step has been described in the above embodiment and will not be repeated in this embodiment.

[0087] As an optional but non-limiting implementation method, the prompt text is input into a large language model to obtain product copy information, including steps C1-C2:

[0088] Step C1: Using a large language model, adjust the keyword weight of the product attribute information according to the copy attribute information;

[0089] Step C2: Determine the product copy information through the large language model according to the copy attribute information and the product attribute information after keyword weight adjustment.

[0090] In this embodiment, in the initial state, each keyword in the text attribute information has a weight, and the weight can be determined based on the evaluation of the important attributes of the product. For example, the initial weights of each text attribute information can be the same. Furthermore, the keyword weights of the product attribute information are adjusted according to the text attribute information through a large language model. For example, for suit-style clothing, the weights of keywords such as "business" and "high-end" will be higher, and for T-shirt-style clothing, the weights of keywords such as "young" and "pattern" will be higher. Under different keywords, the weights of product attribute information are different, which means that one or several items of product attribute information that you want to highlight are different, which is more in line with user expectations and satisfaction.

[0091] Furthermore, the product copy information is determined through a large language model according to the copy attribute information and the product attribute information after keyword weight adjustment.

[0092] For example, for suit-style clothing, if the keywords provided by the copy attribute information are business and high-end, and the copy style is mainly fresh and concise, then through the large language model, according to the product attribute information corresponding to the keywords and keyword weights, you can get a beautiful expression style and appropriate content modification, which constitutes the product copy information of the suit to attract the attention of the target audience. For example, the product copy information of suit-style clothing is "To create a high-end business image, this gray flat-lapel suit is your best choice. It is made of pure wool worsted fabric and combined with a modern design style, which is perfect for all kinds of business meetings and high-end occasions. Whether in the office or in a meeting, this suit can enhance your professional image, allowing you to respond confidently and become the focus of attention."

[0093] In addition, in this embodiment, the product copy information obtained through the large language model can also be processed twice, using natural language processing technology to perform syntax analysis and semantic analysis on the initial product copy information and provide optimization suggestions.

[0094] In addition, feedback and edits on the copywriting are collected and used to optimize the model generation strategy. The model parameters or generation strategy are adjusted based on the feedback to make it more in line with user expectations.

[0095] S260: Use the basic attribute information of the commodity and the commodity text information as commodity library information matching the commodity to be processed.

[0096] It should be noted that this step has been described in the above embodiment and will not be repeated in this embodiment.

[0097] As an optional but non-limiting implementation, the method further includes:

[0098] Generate product display information based on the product images to be processed through the visual big model;

[0099] The basic attribute information of the commodity and the commodity text information are used as commodity library information matching the commodity to be processed, including:

[0100] The basic attribute information of the commodity, the commodity text information and the commodity display information are used as commodity library information matching the commodity to be processed.

[0101] In this embodiment, the visual big model is a deep learning model that can process and understand visual information. Commonly used visual big models include visual big models based on multimodal fusion architecture and visual big models based on encoder-decoder architecture.

[0102] The product display information may be an image or video of the product, and its display form may be 2D or 3D display, which is not specifically limited in this embodiment.

[0103] In this embodiment, the visual big model mainly completes the product modeling and product display. Taking a suit as an example, a deep neural network can be used to convert the 2D image of the product to be processed into a 3D model of the product to be processed, and then a 3D modeling tool can be used to create a customizable virtual model, and the generated 3D suit model is mapped to the virtual model, and the posture and angle of the suit are adjusted. By setting the model's movements and postures, the dynamic effect of the suit on the body is displayed. For this type of information, the 3D display effect of the product can be recorded into an animation based on the display order set in advance to meet the needs of fast display, and real-time 3D model rendering and interaction can be realized on the front end, providing operations such as rotation and scaling to meet the user's needs to view the display effect of the suit from different angles.

[0104] In addition, you can also choose to display only a 2D image of basic attribute information of a product, such as color, texture, style, etc., or display the usage scenario or description of the product, etc. This embodiment is not specifically limited.

[0105] Furthermore, based on the basic attribute information of the product, the product copy information and the product display information, the product library information matching the product to be processed can be formed and stored. The basic attribute information of the product, the input information, the product copy information and the animation can be stored.

[0106] First, the extracted text information is standardized to ensure that information from different sources has a consistent format and structure. The extracted text information is then classified and stored according to the purpose of use. Multiple data warehouses together form a complete system-level non-standard product database. For example, the vector database processing method is used to meet the purpose of image search, and the database storage method is used to store non-standard products and basic attribute information of products, and the relationship between non-standard products. Then, according to the access frequency and modification frequency of the data, the data is cached in layers to balance the cache size and expiration time. Finally, considering the amount of stored data, the overall service expansion is used to disperse the query bottleneck.

[0107] In addition, you can choose to share data to improve the effectiveness of different product references, build a model to calculate user similarity, and provide data sharing services based on user similarity. User similarity will be calculated based on the industry, region, product category composition, customer shopping habits, etc., and substituted into the scoring model.

[0108] See also Figure 3 The figure shows a flow chart of building product library information. The image of the product to be processed is collected by taking photos or recording videos, and the basic attribute information of the product to be processed, such as color, specification and style, is determined based on the image information and text information by obtaining the input information of the user interaction interface or recognizing the text information of the product to be processed by characters. Then, the input information of the user interaction interface is processed to obtain the copy attribute information. According to the basic attribute information of the product and the copy attribute information, the prompt text of the large language model is obtained, and the prompt text is input into the large language model to obtain the product copy information. In addition, through the visual large model, according to the image of the product to be processed, a 2D, 3D rendering or animation of the product display information is generated. Furthermore, based on the product copy information, the basic attribute information of the product and the product display information, it is stored as the product library information of the product to be processed.

[0109] See also Figure 4 The figure shows a module diagram of a product library information. The product library information includes basic product attribute information, product copy information and product display information. The basic product attribute information includes color and style, etc. The product copy information represents the promotional description of the product. The product display information is mainly in the form of display pictures and animations. The product library information can be modified in response to the user interaction interface. The modified product library information result is stored in the product library to reversely train the deep learning model.

[0110] The technical solution of the present application processes the input information through a prompt learner to obtain text attribute information, determines the prompt text according to the basic attribute information of the product and the text attribute information, inputs the prompt text into a large language model to obtain the product text information, and determines the product library information according to the basic attribute information of the product and the product text information, thereby automatically generating the product library information corresponding to the product, reducing the maintenance cost of the product library and the difficulty of product entry, and improving the efficiency of product entry.

[0111] Embodiment 3

[0112] Figure 5 This is a schematic diagram of the structure of a device for generating commodity library information provided in the third embodiment of the present invention. This embodiment is applicable to the generation of commodity library information. The device for generating commodity library information can be implemented in the form of hardware and / or software. The device for generating commodity library information can be configured in any electronic device with network communication and computing capabilities. Figure 5 As shown, the device comprises:

[0113] The commodity attribute information determination module 310 is used to determine the commodity basic attribute information of the commodity to be processed;

[0114] The product copy information determination module 320 is used to determine the prompt text of the large language model according to the basic attribute information of the product and the copy attribute information, and input the prompt text into the large language model to obtain the product copy information;

[0115] The commodity library information determination module 330 is used to use the commodity basic attribute information and the commodity copy information as commodity library information matching the commodity to be processed.

[0116] Optionally, the commodity attribute information determination module 310 includes:

[0117] Determine the product image to be processed;

[0118] Perform image recognition on the commodity image to be processed to obtain at least one feature data;

[0119] The basic attribute information of the commodity to be processed is determined according to the mapping relationship between the feature data and the preset feature data and commodity attributes.

[0120] Optionally, the commodity attribute information determination module 310 includes:

[0121] Determine the product image to be processed;

[0122] Perform text recognition on the product image to be processed to obtain at least one item of text information;

[0123] The text information is converted into structured data as basic attribute information of the commodity to be processed.

[0124] Optionally, the commodity attribute information determination module 310 includes:

[0125] Inputting the basic attribute information of the commodity into the large language model to obtain the attribute information of the commodity after processing;

[0126] The basic attribute information of the commodity and the commodity text information are used as commodity library information matching the commodity to be processed, including:

[0127] The basic attribute information of the commodity, the attribute information of the commodity after processing and the commodity copy information are used as commodity library information matching the commodity to be processed.

[0128] Optionally, the product copy information determination module 320 includes:

[0129] Determine input information for the user interface;

[0130] By prompting the learner, the input information is processed to obtain the copy attribute information;

[0131] According to the basic attribute information of the commodity and the attribute information of the copy, a prompt text of the large language model is determined.

[0132] Optionally, the product copy information determination module 320 includes:

[0133] Through the large language model, the keyword weight of the product attribute information is adjusted according to the copy attribute information;

[0134] Through the large language model, the product copy information is determined based on the copy attribute information and the product attribute information after keyword weight adjustment.

[0135] Optionally, the method further includes:

[0136] Generate product display information based on the product images to be processed through the visual big model;

[0137] The basic attribute information of the commodity and the commodity text information are used as commodity library information matching the commodity to be processed, including:

[0138] The basic attribute information of the commodity, the commodity text information and the commodity display information are used as commodity library information matching the commodity to be processed.

[0139] The technical solution of the present application obtains product copy information through a large language model through product basic attribute information and product copy attribute information, and determines product library information based on the product basic attribute information and product copy information, thereby automatically generating product library information corresponding to the product, reducing product library maintenance costs and product entry difficulty, and improving product entry efficiency.

[0140] The device for generating commodity library information provided in the embodiment of the present invention can execute the method for generating commodity library information provided in any embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method.

[0141] Embodiment 4

[0142] Figure 6 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0143] like Figure 6 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0144] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0145] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a method for generating commodity library information.

[0146] In some embodiments, the method for generating the commodity library information may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method for generating the commodity library information described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the method for generating the commodity library information in any other appropriate manner (e.g., by means of firmware).

[0147] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0148] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0149] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0150] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0151] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0152] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0153] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0154] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for generating commodity library information, characterized in that: include: Determine the basic attribute information of the goods to be processed; Determine a prompt text of a large language model according to the basic attribute information of the product and the attribute information of the copy, and input the prompt text into the large language model to obtain the product copy information; The basic attribute information of the commodity and the commodity text information are used as commodity library information matching the commodity to be processed.

2. The method according to claim 1, characterized in that Determine the basic attribute information of the goods to be processed, including: Determine the product image to be processed; Perform image recognition on the commodity image to be processed to obtain at least one feature data; The basic attribute information of the commodity to be processed is determined according to the mapping relationship between the feature data and the preset feature data and commodity attributes.

3. The method according to claim 1, characterized in that Determine the basic attribute information of the goods to be processed, including: Determine the product image to be processed; Perform text recognition on the product image to be processed to obtain at least one item of text information; The text information is converted into structured data as basic attribute information of the commodity to be processed.

4. The method according to any one of claims 1 to 3, characterized in that: After determining the basic attribute information of the goods to be processed, it also includes: Inputting the basic attribute information of the commodity into the large language model to obtain the attribute information of the commodity after processing; The basic attribute information of the commodity and the commodity text information are used as commodity library information matching the commodity to be processed, including: The basic attribute information of the commodity, the attribute information of the commodity after processing and the commodity copy information are used as commodity library information matching the commodity to be processed.

5. The method according to claim 1, characterized in that According to the basic attribute information of the product and the attribute information of the copy, a prompt text of the large language model is determined, including: Determine input information for the user interface; By prompting the learner, the input information is processed to obtain the copy attribute information; According to the basic attribute information of the commodity and the attribute information of the copy, a prompt text of the large language model is determined.

6. The method according to claim 1, characterized in that The prompt text is input into the large language model to obtain product copy information, including: Through the large language model, the keyword weight of the product attribute information is adjusted according to the copy attribute information; Through the large language model, the product copy information is determined based on the copy attribute information and the product attribute information after keyword weight adjustment.

7. The method according to claim 1, characterized in that The method further comprises: Generate product display information based on the product images to be processed through the visual big model; The basic attribute information of the commodity and the commodity text information are used as commodity library information matching the commodity to be processed, including: The basic attribute information of the commodity, the commodity text information and the commodity display information are used as commodity library information matching the commodity to be processed.

8. A device for generating commodity library information, characterized in that: include: A commodity attribute information determination module, used to determine the commodity basic attribute information of the commodity to be processed; A commodity copy information determination module, used to determine the prompt text of the large language model according to the commodity basic attribute information and the copy attribute information, and input the prompt text into the large language model to obtain the commodity copy information; The commodity library information determination module is used to use the commodity basic attribute information and the commodity copy information as commodity library information matching the commodity to be processed.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method for generating commodity library information as described in any one of claims 1 to 7 is implemented.

10. A storage medium storing computer executable instructions, characterized in that: The computer executable instructions are used to execute the method for generating commodity library information as described in any one of claims 1 to 7 when executed by a computer processor.

Images