Product business document generation method and device based on large model and related equipment

Abstract

The invention provides a product business document generation method and device based on a large model and related equipment, and relates to the technical field of artificial intelligence. The method comprises the following steps: acquiring product business description information of a target product; according to the product business description information, determining each directory chapter included in the to-be-generated product business document; according to each catalogue chapter included in the to-be-generated product business document, chapter content included in each catalogue chapter is generated by utilizing a pre-trained product large model and combining a pre-constructed product business specification vector library and a product knowledge graph, the product business specification vector library comprises product business specification information of a plurality of stock products; the product knowledge graph comprises product total information of a plurality of stock products; and generating a product business document of the target product according to chapter content included in each directory chapter. According to the method and the device, the product business document meeting requirements can be quickly and accurately generated.

Description

Background Art

[0002] For communication operators, before new products are launched, the group marketing department is often required to issue product business specifications of new products to the entire network (usually including product descriptions, product finalization design, product network security, etc.). At present, product business specifications are jointly written by the marketing department, IT department, security department, customer service department and other relevant departments after discussion. When the operator's large network is about to launch new products from some professional companies, since the product managers of the professional companies are not familiar with the content of the product business specifications of the large network system, it often takes a lot of manpower to complete a relevant document of product business specifications, which is time-consuming.

[0003] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0004] The present disclosure provides a method for generating a product business document based on a large model and related equipment, which at least to a certain extent overcomes the technical problems in the related art of manually writing product business documents, which have high labor costs and long time consumption.

[0005] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0006] According to one aspect of the present disclosure, a method for generating a product business document based on a big model is provided, including: obtaining product business description information of a target product; determining, based on the product business description information, each catalog chapter included in the product business document to be generated; based on the each catalog chapter included in the product business document to be generated, using a pre-trained product big model, combined with a pre-built product business specification vector library and a product knowledge graph, generating chapter content included in each catalog chapter, wherein the product business specification vector library includes: product business specification information of multiple stock products; the product knowledge graph includes: full product information of multiple stock products; generating the product business document of the target product based on the chapter content included in each catalog chapter.

[0007] In some embodiments, according to each catalog chapter of the product business document to be generated, a pre-trained product big model is used, combined with a pre-built product business specification vector library and a product knowledge graph, to generate chapter content contained in each catalog chapter, including: according to the product business description information of the target product, searching for the product business specification information of the target product from the pre-built product business specification vector library; according to each catalog chapter contained in the product business document to be generated, a pre-trained product big model is used to conduct multiple question-and-answer dialogues with the target user, and according to the dialogue content output by the product big model in each question-and-answer dialogue, a product knowledge graph subgraph of the target product is dynamically constructed, wherein the product knowledge graph subgraph is used to obtain product association information of the target product; the product business specification information of the target product and the product knowledge graph subgraph are fused to obtain the fusion information of the target product; each catalog chapter contained in the product business document to be generated, the fusion information of the target product and the preset prompt words are input into the product big model, and the chapter content contained in each catalog chapter in the product business document to be generated is output.

[0008] In some embodiments, based on the various catalog chapters contained in the product business document to be generated, a pre-trained product big model is used to conduct multiple question-and-answer dialogues with the target user, and based on the dialogue content output by the product big model in each question-and-answer dialogue, a product knowledge graph subgraph of the target product is dynamically constructed, including: using the dialogue content output by the product big model in each question-and-answer dialogue as a node, searching for product-related information from the product knowledge graph, and generating a corresponding product knowledge graph subgraph; and updating the product knowledge graph subgraph corresponding to the dialogue content output by the product big model in historical question-and-answer dialogues.

[0009] In some embodiments, before generating chapter contents contained in each catalog chapter according to each catalog chapter contained in the product business document to be generated, using a pre-trained product big model, combined with a pre-built product business specification vector library and a product knowledge graph, the method also includes: obtaining product private domain data; and fine-tuning a pre-trained big language model according to the product private domain data to obtain the product big model.

[0010] In some embodiments, before generating chapter contents contained in each catalog chapter according to each catalog chapter contained in the product business document to be generated, using a pre-trained product big model in combination with a pre-built product business specification vector library and a product knowledge graph, the method further includes: obtaining product business specification information of multiple stock products; and constructing a product business specification vector library based on the product business specification information of the multiple stock products.

[0011] In some embodiments, before generating chapter contents contained in each catalog chapter according to each catalog chapter contained in the product business document to be generated, using a pre-trained product big model in combination with a pre-built product business specification vector library and a product knowledge graph, the method also includes: obtaining full product information of multiple stock products; and constructing a product knowledge graph based on the full product information of the multiple stock products.

[0012] In some embodiments, after generating chapter contents contained in each catalog chapter according to each catalog chapter contained in the product business document to be generated, using a pre-trained product big model in combination with a pre-built product business specification vector library and a product knowledge graph, the method further includes: displaying the chapter contents contained in each catalog chapter to the target user; receiving editing operations of the target user on the chapter contents under one or more target chapters; and adjusting the chapter contents under the corresponding target chapter according to the editing operations.

[0013] In some embodiments, the product business document is a product business specification document.

[0014] According to one aspect of the present disclosure, a device for generating a product business document based on a big model is also provided, including: an information collection module, used to obtain product business description information of a target product; a product business document directory chapter determination module, used to determine the directory chapters contained in the product business document to be generated according to the product business description information; a product business document intelligent writing module, used to generate the chapter content contained in each directory chapter according to the directory chapters contained in the product business document to be generated, using a pre-trained product big model, combined with a pre-built product business specification vector library and a product knowledge graph, wherein the product business specification vector library contains: product business specification information of multiple stock products; the product knowledge graph contains: full product information of multiple stock products; a product business document generation module, used to generate the product business document of the target product according to the chapter content contained in each directory chapter.

[0015] According to another aspect of the present disclosure, an electronic device is also provided, including: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute any one of the above-mentioned methods for generating product business documents based on a large model by executing the executable instructions.

[0016] According to another aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method for generating a product business document based on a large model described above is implemented.

[0017] According to another aspect of the present disclosure, a computer program product is provided, including: a computer program or instructions, which, when executed by a processor, implements any one of the above-mentioned methods for generating product business documents based on a large model.

[0018] The method, device and related equipment for generating a product business document based on a big model provided in the embodiments of the present disclosure pre-build a product business specification vector library containing multiple product business specification information and a product knowledge graph containing multiple product full information, and use the product private domain data to fine-tune the big language model to obtain a product big model, so as to use the trained product big model, combined with the pre-built product business specification vector library and product knowledge graph, when generating a product business document of a certain product, it is only necessary to obtain the product business description information of the product, so as to determine the various directory chapters contained in the product business document to be generated, and generate the chapter contents contained in each directory chapter, and finally generate the product business document of the product according to the chapter contents contained in each directory chapter.

[0019] Through the embodiments of the present disclosure, it is possible to quickly and accurately generate product business documents that meet the needs, thereby solving the problem of high cost and long time consumption caused by manual writing of product business documents.

[0020] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without creative work.

[0022] Figure 1 A schematic diagram of an application system architecture in an embodiment of the present disclosure is shown;

[0023] Figure 2 A flow chart of a method for generating a product business document based on a large model in an embodiment of the present disclosure is shown;

[0024] Figure 3 A flowchart of writing chapter content based on a large model in an embodiment of the present disclosure is shown;

[0025] Figure 4 A flow chart of a method for dynamically constructing a product knowledge graph subgraph in an embodiment of the present disclosure is shown;

[0026] Figure 5 A flow chart of a method for generating a large product model in an embodiment of the present disclosure is shown;

[0027] Figure 6 A flow chart of a method for constructing a product business specification vector library in an embodiment of the present disclosure is shown;

[0028] Figure 7 A flow chart of a method for constructing a product knowledge graph in an embodiment of the present disclosure is shown;

[0029] Figure 8 A flow chart of a method for manually editing, updating and adjusting chapter content generated by a product macro model in an embodiment of the present disclosure is shown;

[0030] Fig. 9 A flowchart of a method for manually writing product business documents in the related art is shown;

[0031] Fig.10 A flow chart of a method for intelligently writing product business documents based on a large model in an embodiment of the present disclosure is shown;

[0032] Fig.11 A schematic diagram of a device for generating product business documents based on a large model in an embodiment of the present disclosure is shown;

[0033] Fig.12 A structural block diagram of an electronic device in an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0034] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the disclosure will be more comprehensive and complete and to fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0035] In addition, the accompanying drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0036] The specific implementation of the embodiment of the present disclosure is described in detail below with reference to the accompanying drawings.

[0037] Figure 1FIG. 1 shows an exemplary application system architecture diagram to which the method for generating product business documents based on a large model in an embodiment of the present disclosure can be applied. Figure 1 As shown, the system architecture may include a terminal device 10 and a server 30 .

[0038] The network 20 is a medium for providing a communication link between the terminal device 10 and the server 30, and may be a wired network or a wireless network.

[0039] Optionally, the wireless network or wired network described above uses standard communication technology and / or protocol. The network is usually the Internet, but it can also be any network, including but not limited to a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a dedicated network or any combination of a virtual private network). In some embodiments, the data exchanged through the network is represented by technologies and / or formats including Hyper Text Mark-up Language (HTML), Extensible Markup Language (XML), etc. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPSec) can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above data communication technologies.

[0040] The terminal device 10 can be various electronic devices, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, smart speakers, smart watches, wearable devices, augmented reality devices, virtual reality devices, etc.

[0041] Optionally, the client of the application installed in different terminal devices 10 is the same, or the client of the same type of application based on different operating systems. Based on the different terminal platforms, the specific form of the client of the application can also be different, for example, the application client can be a mobile client, a PC client, etc.

[0042] The server 30 may be a server that provides various services, such as a background management server that provides support for the device operated by the user using the terminal device 10. The background management server may analyze and process the received request and other data, and feed back the processing results to the terminal device.

[0043] Optionally, the server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as big data and artificial intelligence platforms.

[0044] Those skilled in the art will know that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration, and any number of terminal devices, networks and servers may be provided according to actual needs, and the embodiments of the present disclosure do not limit this.

[0045] Under the above system architecture, a method for generating product business documents based on a large model is provided in an embodiment of the present disclosure. The method can be executed by any electronic device with computing and processing capabilities.

[0046] In some embodiments, the method for generating product business documents based on big models provided in the embodiments of the present disclosure can be executed by a terminal device of the above-mentioned system architecture; in other embodiments, the method for generating product business documents based on big models provided in the embodiments of the present disclosure can be executed by a server in the above-mentioned system architecture; in other embodiments, the method for generating product business documents based on big models provided in the embodiments of the present disclosure can be implemented by the terminal device and server in the above-mentioned system architecture through interaction.

[0047] Figure 2 A flowchart of a method for generating a product business document based on a large model in an embodiment of the present disclosure is shown. Figure 2 As shown, the method for generating a product business document based on a large model provided in an embodiment of the present disclosure includes the following steps:

[0048] S202, obtaining product business description information of the target product.

[0049] It should be noted that the target product can be any product, including software products and physical products. In the embodiments of the present disclosure, the products released by the operator are used as examples to illustrate various embodiments. The product service description information can be the description information of some products and services required for generating the catalog chapters of the product service documents related to the target product. Different product service documents may contain different catalogs, and therefore different product service description information may need to be obtained.

[0050] In one embodiment, when the target product is an operator's product and the product business document to be generated is a product business specification document, the product business description information obtained in S202 may be basic information of the target product (such as product name, etc.), product application basis, and product finalization information.

[0051] In the embodiment of the present disclosure, the product business description information obtained in the above S202 can be directly input by the user, or can be obtained from a pre-built product database according to the product name or logo. The present disclosure does not limit the method of obtaining the product business description information and the specific content obtained, and aims to protect a method that can intelligently write relevant product business documents based on a large model according to the product business description information.

[0052] S204: Determine the catalog chapters included in the product business document to be generated according to the product business description information.

[0053] It should be noted that the product service document to be generated may be any product service document. In the embodiments of the present disclosure, the product service specification document is taken as an example to illustrate each embodiment.

[0054] S206, according to each catalog chapter included in the product business document to be generated, using the pre-trained product big model, combined with the pre-built product business specification vector library and product knowledge graph, generate the chapter content included in each catalog chapter, wherein the product business specification vector library includes: product business specification information of multiple stock products; the product knowledge graph includes: full product information of multiple stock products;

[0055] It should be noted that the above-mentioned product big model is a big model obtained by fine-tuning a pre-trained big language model according to the product private domain data, and can output dialogue content that conforms to human question-and-answer habits. The above-mentioned product name business specification vector library is a pre-built database containing product business specification information of multiple stock products; the above-mentioned product knowledge graph is a pre-built knowledge graph containing the full product information of multiple stock products. In the embodiment of the present disclosure, stock products refer to products whose full product information and product business specification information are known.

[0056] After determining the various directory chapters included in the product business document to be generated, the pre-trained product big model is used, combined with the pre-built product business specification vector library and product knowledge graph, and multiple questions and answers are conducted with the target user to generate the chapter content contained in each directory chapter in the product business document to be generated. In the specific implementation, according to the various directory chapters included in the product business document to be generated, according to the predetermined document structure, through a series of question-and-answer dialogues, the model is guided to intelligently generate a product business document that meets the specifications. The disclosed embodiment uses a big model, a vector library, and a knowledge graph to understand and predict the specific needs or intentions of business personnel, and can solve the cold start problem (i.e., the difficulty of starting to generate documents without sufficient prior information) when generating product business specification information and the hallucination problem of the big model (i.e., the inaccurate or illogical content generated by the model).

[0057] In one embodiment, when generating the chapter contents contained in each catalog chapter in the product business document to be generated, it is also possible to combine the customized prompts of each catalog chapter, use the pre-trained product big model, conduct a question-and-answer dialogue with the target user, and combine the pre-built product business specification vector library and product knowledge graph to generate the chapter contents contained in each catalog chapter. In specific implementation, the prompts and the information that has undergone knowledge fusion can be input into the product big model together, and the product big model returns the latest question and answer content for document generation. In the disclosed embodiment, the prompts and the information that has undergone knowledge fusion are input into the big model together, and the big model returns the latest question and answer content for document generation.

[0058] Furthermore, in one embodiment, the writing requirements of the product business document can be analyzed, and customized prompts for different catalog chapters can be automatically generated. Then, according to the feedback content of the product macro model, the corresponding prompts can be automatically adjusted to improve the accuracy and relevance of the document generation.

[0059] S208, generating a product business document of the target product according to the chapter contents contained in each catalog chapter.

[0060] It should be noted that after the chapter contents contained in each catalog chapter in the to-be-generated product business document are generated through the product macro model, the corresponding product business document can be generated according to each catalog chapter of the to-be-generated product business document and the chapter contents contained in each catalog chapter. The product business document generated in the embodiment of the present disclosure can be a document in any preset format. In one embodiment, it can be an online product business document in WORD format.

[0061] In one embodiment, a large model is used to optimize and adjust the language of the generated document to ensure that the generated document meets the specific needs of business personnel.

[0062] In some embodiments, the product business document in the embodiments of the present disclosure is a product business specification document (also referred to as a product business specification publication) that publishes product business specification information.

[0063] From the above, it can be seen that the method for generating product business documents based on a big model provided in the embodiment of the present disclosure pre-constructs a product business specification vector library containing multiple product business specification information and a product knowledge graph containing full information of multiple products, and uses the product private domain data to fine-tune the big language model to obtain a product big model, so as to utilize the trained product big model, combined with the pre-constructed product business specification vector library and product knowledge graph, when generating a product business document for a certain product, it is only necessary to obtain the product business description information of the product, so as to determine the various directory chapters contained in the product business document to be generated, and generate the chapter contents contained in each directory chapter, and finally generate the product business document of the product according to the chapter contents contained in each directory chapter.

[0064] Through the large model-based product business document generation method provided in the embodiments of the present disclosure, product business documents that meet the needs can be generated quickly and accurately, solving the problem of high cost and long time consumption caused by manual writing of product business documents.

[0065] In some embodiments, Figure 3 As shown, the method for generating product business documents based on a large model provided in the embodiment of the present disclosure can

[0066] S302, searching for product business specification information of the target product from a pre-built product business specification vector library according to the product business description information of the target product;

[0067] S304: Based on the various catalog chapters contained in the product business document to be generated, the pre-trained product big model is used to conduct multiple question-and-answer dialogues with the target user, and based on the dialogue content output by the product big model in each question-and-answer dialogue, a product knowledge graph subgraph of the target product is dynamically constructed, wherein the product knowledge graph subgraph is used to obtain product association information of the target product;

[0068] S306, fusing the product business specification information of the target product with the product knowledge graph subgraph to obtain fusion information of the target product;

[0069] S308, inputting each catalog chapter, fusion information of the target product and preset prompt words contained in the product business document to be generated into the product macro model, and outputting the chapter content contained in each catalog chapter in the product business document to be generated.

[0070] In some embodiments, Figure 4As shown, the method for generating product business documents based on a large model provided in the embodiment of the present disclosure can dynamically construct a product knowledge graph subgraph of the target product through the following steps:

[0071] S402, using the dialogue content output by the product big model in each question-and-answer dialogue as a node, searching for product related information from the product knowledge graph, and generating a corresponding product knowledge graph subgraph;

[0072] S404, based on the product knowledge graph subgraph corresponding to the dialogue content output by the product macro model during the historical question-and-answer dialogue, update the product knowledge graph subgraph corresponding to the dialogue content output during the current question-and-answer dialogue.

[0073] It should be noted that during the document generation process, the answers returned by the large model are recorded as nodes, and the information related to these answers is inferred. When generating new answer content, relevant information is extracted from the product knowledge graph, and combined with the product knowledge graph subgraph corresponding to the historical answer content, the relevance of the content is calculated, irrelevant nodes are deleted, and the product knowledge graph subgraph is dynamically updated to obtain a connected subgraph corresponding to the new answer content.

[0074] In the disclosed embodiment, based on the reasoning ability of the past large model question and answer and knowledge graph, the product knowledge graph subgraph is dynamically constructed in real time. The subgraph is more related to document generation, the content quality is improved, the computing overhead is reduced, and effective information can be obtained efficiently. The product knowledge graph subgraph is combined with the product business posting vector library to provide more information as input for the product large model, effectively solve the cold start problem, and reduce the illusion of the large model.

[0075] In some embodiments, Figure 5 As shown, the method for generating a product business document based on a big model provided in the embodiment of the present disclosure can generate a product big model through the following steps:

[0076] S502, obtaining product private domain data;

[0077] S504: fine-tune the pre-trained large language model according to the product private domain data to obtain a large product model.

[0078] In the disclosed embodiment, a product big model is trained based on stock product data to improve its generalization ability. The product big model is combined with the product knowledge graph and vector library to intelligently generate product business specification documents that meet user requirements according to the document directory based on user needs. Fine-tuning the big model based on product private domain knowledge can obtain a product big model that is more suitable for a specific product field.

[0079] Figure 6 A flow chart of a method for constructing a product business specification vector library in an embodiment of the present disclosure is shown as follows: Figure 6 As shown, in some embodiments, before generating the chapter contents contained in each catalog chapter according to each catalog chapter contained in the product business document to be generated, using a pre-trained product big model in combination with a pre-built product business specification vector library and a product knowledge graph, the product business document generation method based on the big model provided in the embodiment of the present disclosure may further include the following steps:

[0080] S602, obtaining product business specification information of multiple stock products;

[0081] S604: construct a product business specification vector library based on the product business specification information of multiple stock products.

[0082] In specific implementation, a corresponding vector representation can be generated according to the product business specification information of each stock product, and then a product business specification vector library can be constructed according to the product business specification vector representations of multiple stock products and the corresponding product business description information. After obtaining the product business description information of the target product, the product business specification information of similar products can be searched from the product business specification vector library according to the product business description information of the target product. In the embodiment of the present disclosure, similar products refer to products whose similarity of product business description information meets a preset condition (for example, greater than a preset threshold).

[0083] In the embodiments of the present disclosure, by constructing a product business specification vector library, the purpose of quickly searching and matching product business specification information can be achieved.

[0084] Figure 7 A flow chart of a method for constructing a product knowledge graph in an embodiment of the present disclosure is shown as follows: Figure 7 As shown, in some embodiments, before generating the chapter contents contained in each catalog chapter according to each catalog chapter contained in the product business document to be generated, using a pre-trained product big model in combination with a pre-built product business specification vector library and a product knowledge graph, the product business document generation method based on the big model provided in the embodiment of the present disclosure may further include the following steps:

[0085] S702, obtaining full product quantity information of multiple stock products;

[0086] S704: Build a product knowledge graph based on the full product information of multiple stock products.

[0087] It should be noted that the above-mentioned full product information includes but is not limited to product business description information, product tariffs, product attributes, business specifications, processes, finalization, product relationships, etc.; the above-mentioned product knowledge graph can be a product knowledge graph involving full information such as product business description information, product tariffs, product attributes, business specifications, processes, finalization, product relationships, etc. When the product business description information (such as description, function type) of the target product is input, the product finalization, process, attribute and other information associated with it are inferred through the constructed product knowledge graph.

[0088] In one embodiment, the big model-based product document generation method provided in the embodiments of the present disclosure may also include the following steps: regularly collecting product-related information from external data sources; automatically identifying and integrating new product information into the existing knowledge graph; and optimizing the structure of the knowledge graph to improve query efficiency and accuracy.

[0089] In one embodiment, the product document generation method based on the big model provided in the embodiment of the present disclosure may also include the following steps: using the big model to conduct a preliminary review of the generated document content; identifying and correcting errors, inconsistencies or missing information in the document; and automatically making improvement suggestions to optimize the document quality.

[0090] In one embodiment, the large model-based product document generation method provided in the embodiments of the present disclosure may also include the following steps: receiving product-related information input by a user; displaying the progress and status of document generation; and allowing the user to provide real-time feedback and modification to the generated document content.

[0091] Figure 8 A flow chart of a method for manually editing, updating and adjusting chapter content generated by a product macro model according to an embodiment of the present disclosure is shown. Figure 8 As shown, in some embodiments, after generating the chapter contents contained in each catalog chapter according to each catalog chapter contained in the product business document to be generated, using the pre-trained product big model in combination with the pre-built product business specification vector library and product knowledge graph, the product business document generation method based on the big model provided in the embodiment of the present disclosure can also manually edit, update and adjust the chapter contents generated by the product big model through the following steps:

[0092] S802, displaying the chapter contents contained in each directory chapter to the target user;

[0093] S804, receiving an editing operation of a target user on the content of one or more target chapters;

[0094] S806: Adjust the chapter content under the corresponding target chapter according to the editing operation.

[0095] When it is necessary to explain, after the chapter contents contained in each catalog chapter output by the product big model are integrated into the relevant product business document according to the catalog chapters of the product document to be generated, it can be returned to the user for online editing to adjust the chapter contents under the corresponding target chapter in the product business document.

[0096] In one embodiment, the method for generating product business documents based on a large model provided in the embodiment of the present disclosure may further include the following steps: monitoring the performance indicators of the product large model during the document generation process; adjusting the model parameters according to the performance data. Through this embodiment, the document generation efficiency of the product large model can be improved, and dynamic resource allocation can be achieved to meet different document generation requirements.

[0097] Fig. 9 A schematic diagram of a product business document production system architecture in an embodiment of the present disclosure is shown. Fig. 9 As shown, when using the product, the user only needs to collect the basic product information, product application basis, and product finalization information. The document intelligent writing module can intelligently analyze the content that needs to be written for the product based on the information input by the user, existing knowledge, context, etc., and automatically complete the content under the product implementation, customer service management, security management, etc. The document generation module generates product business specification documents based on catalog integration and feeds back to the user, supporting online editing and adjustment. Compared with the existing product business document system, the implementation of the disclosed embodiment only requires optimizing the information collection module and adding a document intelligent writing module.

[0098] Taking the product business document as the product business specification document as an example, the method provided in the embodiment of the present disclosure can be based on the existing product business specification documents, train the private domain product big model, combine the product business specification vector library and the product knowledge graph, and intelligently generate product business specifications according to the document directory chapters through multiple questions and answers. The embodiment of the present disclosure, through the product big model combined with the vector library and the product knowledge graph, identifies the intention of the business personnel, solves the problem of cold start of product business specification generation, and reduces the appearance of big model illusions. Furthermore, based on the big model, the document is polished and adjusted to generate a business document that meets the needs of the business personnel.

[0099] It should be noted that the relevant technologies rely on the method of manually generating product business specification documents to collect information such as product description, product implementation, customer service management, security management, product delivery, etc. When collecting this information, it requires the cooperation of multiple professional personnel from different departments, and it takes a long time to collect a large amount of information; and the product managers of third-party professional companies may not be familiar with the relevant rules of the big network, and the cost of offline manual communication is high.

[0100] Fig.10 FIG. 1 is a schematic diagram showing an overall framework of a product business document intelligent writing module according to an embodiment of the present disclosure. Fig.10 As shown, the product business document intelligent writing module specifically includes: information input module 101, product business specification vector library 102, product knowledge graph 103, product knowledge graph subgraph 104, knowledge fusion module 105, prompt engineering module 106, product large model 107 and model output module 108. The functions implemented by each module are as follows:

[0101] Information input module 101: collect basic information of the product, product application basis, and product finalization information as the basis for writing the product description chapter.

[0102] Product business specification vector library 102: Based on the product stock business specification, a vector library is constructed. When information is input, similar information is searched in combination with the vector library content.

[0103] Product knowledge graph 103: Construct a product knowledge graph involving all information such as basic product information, product prices, product attributes, business specifications, processes, finalization, and relationships between products. When basic product information (such as description, function type) is input, the graph is used to infer the product finalization, process, attribute and other information associated with it.

[0104] Product knowledge graph subgraph 104: records the answers returned by the big model during the document content generation process as nodes, and infers the relevant information associated with its answers. When generating the latest content, obtain the information associated with the content from the product knowledge graph, and combine it with the knowledge subgraph of the answers returned in the past, calculate the relevance of the content, delete irrelevant nodes, dynamically update the knowledge subgraph, and obtain the connected subgraph corresponding to the latest content.

[0105] Knowledge fusion module 105: finds similar information in the product business specification vector library and fuses it with the content of the knowledge subgraph.

[0106] Prompt Engineering Module 106: Pre-make prompts for different scenarios, such as: Please refer to, product: {product_name}, product description: {product_description}. Please list the "usage scenarios" and output them strictly in the format. No summarization is allowed, so that the large model can return the answer as expected.

[0107] Product big model 107: The product big model is obtained after fine-tuning based on the product private domain knowledge, and the prompt words and knowledge fusion information are input into the big model together, and the big model returns the latest question and answer content.

[0108] The product business document generation method provided by the embodiment of the present disclosure is based on a generative big model, according to the user's needs and the reasoning ability of the stock knowledge, automatically writes the corresponding content according to the document directory, and then automatically generates the product business specification document. Through the embodiment of the present disclosure, the problem of document cold start can be solved, and the document content that meets the user's intention can be generated according to the user's needs.

[0109] The method for generating product business documents based on a large model in the embodiments of the present disclosure can achieve but is not limited to the following technical effects:

[0110] ① Build an automatic generation architecture for product business documents based on the information collection module, document intelligent writing module, and document generation module. The information collection module simplifies the collected content; the document intelligent writing module first determines the document directory chapters, and then based on the information input by the user combined with the context content and other information, the product big model intelligently writes the content under the document directory chapters; the document generation module integrates the content feedback from the product big model into product business documents according to the document directory, and returns it to the user, supporting online editing and adjustment by the user. This method enables the intelligent generation of product business documents, reduces the workload of users in writing documents, and increases the speed of product listing.

[0111] ② Combine the product knowledge graph subgraph (product business specification graph) with the product business document vector library to provide more information as input for the product big model, effectively solve the document cold start problem, and reduce the illusion of the big model.

[0112] ③ Based on the reasoning ability of past large-model question-answering and product knowledge graphs, the product knowledge graph subgraph is dynamically updated in real time. When the latest content is generated, the information associated with the content is obtained from the product knowledge graph, and combined with the knowledge subgraph of the past returned answers, the relevance of the content is calculated, irrelevant nodes are deleted, and the knowledge subgraph is dynamically updated to obtain the connected subgraph corresponding to the latest content. The subgraph focuses more on information related to document generation, the content quality is improved, the computing overhead is reduced, and effective information can be obtained efficiently.

[0113] ④ Based on the stock product data, train the product big model to improve its generalization ability. The product big model is combined with the product knowledge graph and product business specification vector library to intelligently generate product business specification documents that meet user requirements according to the document directory based on user needs.

[0114] It should be noted that the existing RAG (Retrieval-augmented Generation) method is to plug the knowledge vector library into the big model, that is, after collecting the basic information of the product in the system, the basic information of the product is combined with the product knowledge vector library, and sent to the big model through the prompt engineering assistance. The big model feedbacks the question and answer content of the product. For example: input the product description and require the big model to output the applicable scenario of the product. Then the existing method will search the knowledge base for applicable scenarios corresponding to similar descriptions based on the input product description, and then input the content into the big model through the prompt engineering. The big model answers the applicable scenario of the product based on the product description and the applicable scenarios of similar products. When continuing to ask questions, if the product usage scenario is entered, the big model is required to output the user group of the product in the scenario. Then the existing method will return the answer through the big model based on the usage scenario and the knowledge base, and will not feedback the answer based on the product description information that has been entered before. Therefore, this method has a weak ability to combine contextual information.

[0115] In the disclosed embodiment, a product knowledge graph and a product knowledge graph subgraph are added. The product knowledge graph subgraph is different from the product knowledge graph. The product knowledge graph subgraph is a product knowledge graph subgraph dynamically constructed based on the question and answer information and the reasoning ability of the product knowledge graph in the process of answering questions to the large model. The subgraph is mainly to make up for the context information lost in the large model question and answer process. As in the above scenario, after asking about the usage scenario under the product description, the user group is asked to the large model. When the product knowledge graph subgraph asks about the product scenario, the product description, product scenario and the information inferred by the product are constructed as nodes of the product knowledge graph subgraph. The subgraph will be continuously updated according to the large model question and answer information, and the nodes in the graph will be dynamically constructed. When the user group in the scenario is asked according to the product usage scenario, the method will obtain the context content (i.e., product description, product usage scenario and other related information) and reasoning information through the subgraph information, and then merge it with the content in the knowledge base and send it to the large model, and the large model will feedback the answer. The dynamic construction of product knowledge graph subgraphs and knowledge fusion in the disclosed embodiment are all missing contents in the RAG method. The disclosed embodiment can better understand contextual information and semantic information, reduce the illusion of large models, make the generated content more coherent and consistent, and avoid ambiguity and misunderstanding.

[0116] The method for generating product business documents based on a big model provided in the embodiments of the present disclosure will intelligently generate an organizational framework for product business document releases during the document generation process, and then generate the content of the document under the organizational framework by combining the big model with a vector library and a product knowledge graph subgraph. The method for intelligently generating an organizational framework is as follows: the document structure is sorted out and precipitated into a template library, and the intent is recognized through a big model based on the document information that the user wants to generate, and then the organizational structure of the corresponding document is matched. For example, if the user inputs "I want a business specification document about government and enterprise products", the big model intent recognition obtains the keywords "government and enterprise products" and "business specification documents", matches them with the template library, and obtains the organizational structure of the business specification documents for government and enterprise products. Subsequent documents are generated according to the organizational structure.

[0117] Based on the same inventive concept, the present disclosure also provides a large model-based product business document generation device, as described in the following embodiments. Since the principle of solving the problem in the device embodiment is similar to that in the above method embodiment, the implementation of the device embodiment can refer to the implementation of the above method embodiment, and the repeated parts will not be repeated.

[0118] Fig.11 A schematic diagram of a device for generating product business documents based on a large model in an embodiment of the present disclosure is shown. Fig.11 As shown, the device includes: an information collection module 111, a product business document catalog chapter determination module 112 and a product business document intelligent writing module 113.

[0119] Among them, the information collection module 111 is used to obtain the product business description information of the target product; the product business document directory chapter determination module 112 is used to determine the various directory chapters included in the product business document to be generated based on the product business description information; the product business document intelligent writing module 113 is used to generate the chapter content contained in each directory chapter based on the various directory chapters included in the product business document to be generated, using a pre-trained product big model, combined with a pre-built product business specification vector library and product knowledge graph, wherein the product business specification vector library contains: product business specification information of multiple existing products; the product knowledge graph contains: full product information of multiple existing products; the product business document generation module is used to generate the product business document of the target product based on the chapter content contained in each directory chapter.

[0120] In some embodiments, the above-mentioned product business document intelligent writing module 113 specifically includes: a product business specification vector library module 1131, a product knowledge graph subgraph dynamic construction module 1132, an information fusion module 1133 and a product large model writing module 1134.

[0121] Among them, the product business specification vector library module 1131 is used to search for the product business specification information of the target product from the pre-built product business specification vector library according to the product business description information of the target product; the product knowledge graph subgraph dynamic construction module 1132 is used to use the pre-trained product big model to conduct multiple question-and-answer dialogues with the target user according to the various directory chapters contained in the product business document to be generated, and dynamically construct the product knowledge graph subgraph of the target product according to the dialogue content output by the product big model in each question-and-answer dialogue, wherein the product knowledge graph subgraph is used to obtain the product-related information of the target product; the information fusion module 1133 is used to fuse the product business specification information of the target product with the product knowledge graph subgraph to obtain the fusion information of the target product; the product big model writing module 1134 is used to input the various directory chapters contained in the product business document to be generated, the fusion information of the target product and the preset prompt words into the product big model, and output the chapter content contained in each directory chapter in the product business document to be generated.

[0122] Furthermore, in some embodiments, the above-mentioned product knowledge graph subgraph dynamic construction module 1132 is also used to: use the dialogue content output by the product big model in each question and answer dialogue as a node, search for product-related information from the product knowledge graph, and generate a corresponding product knowledge graph subgraph; according to the product knowledge graph subgraph corresponding to the dialogue content output by the product big model in historical question and answer dialogues, update the product knowledge graph subgraph corresponding to the dialogue content output in the current question and answer dialogue.

[0123] Furthermore, in some embodiments, the product business document generation device based on the big model provided in the embodiments of the present disclosure may also include: a product big model training module 114, which is used to obtain product private domain data; according to the product private domain data, the pre-trained big language model is fine-tuned to obtain the product big model.

[0124] Furthermore, in some embodiments, the large model-based product business document generation device provided in the embodiments of the present disclosure may also include: a product business specification vector library construction module 115, which is used to obtain product business specification information of multiple stock products; and construct a product business specification vector library based on the product business specification information of multiple stock products.

[0125] In some embodiments, the large model-based product business document generation device provided in the embodiments of the present disclosure may also include: a product knowledge graph construction module 116, which is used to obtain the full product information of multiple stock products; and construct a product knowledge graph based on the full product information of multiple stock products.

[0126] In some embodiments, the large model-based product business document generation device provided in the embodiments of the present disclosure may also include: a user module 117, which is used to display the chapter content contained in each directory chapter to the target user; receive the target user's editing operations on the chapter content under one or more target chapters; and adjust the chapter content under the corresponding target chapter according to the editing operations.

[0127] In some embodiments, the product business document may be, but is not limited to, a product business specification document.

[0128] It should be noted that the examples and application scenarios implemented by the modules in the above-mentioned device embodiment are the same as those of the corresponding steps in the method embodiment, but are not limited to the contents disclosed in the above-mentioned method embodiment. It should be noted that the above-mentioned modules as part of the device can be executed in a computer system such as a set of computer executable instructions.

[0129] Those skilled in the art will appreciate that various aspects of the present disclosure may be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation that combines hardware and software aspects, which may be collectively referred to herein as a "circuit," "module," or "system."

[0130] Based on the same inventive concept, an electronic device is also provided in an embodiment of the present disclosure, the electronic device comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute any one of the above-mentioned methods for generating product business documents based on a large model by executing the executable instructions. Since the principle of solving the problem in the electronic device embodiment is similar to that in the above-mentioned method embodiment, the implementation of the electronic device embodiment can refer to the implementation of the above-mentioned method embodiment, and the repeated parts will not be repeated.

[0131] Refer to the following Fig.12 1200 according to this embodiment of the present disclosure is described. Fig.12 The electronic device 1200 shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0132] like Fig.12 As shown, the electronic device 1200 is in the form of a general computing device. The components of the electronic device 1200 may include but are not limited to: at least one processing unit 1210, at least one storage unit 1220, and a bus 1230 connecting different system components (including the storage unit 1220 and the processing unit 1210).

[0133] The storage unit stores a program code, and the program code can be executed by the processing unit 1210, so that the processing unit 1210 executes the steps described in the above "Exemplary Method" section of this specification according to various exemplary embodiments of the present disclosure. For example, the processing unit 1210 can execute the following steps of the above method embodiment: obtain product business description information of the target product; determine the various catalog chapters included in the product business document to be generated based on the product business description information; based on the various catalog chapters included in the product business document to be generated, use a pre-trained product big model, combined with a pre-built product business specification vector library and a product knowledge graph, to generate the chapter content contained in each catalog chapter, wherein the product business specification vector library contains: product business specification information of multiple stock products; the product knowledge graph contains: full product information of multiple stock products; based on the chapter content contained in each catalog chapter, generate the product business document of the target product.

[0134] The storage unit 1220 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 12201 and / or a cache storage unit 12202 , and may further include a read-only storage unit (ROM) 12203 .

[0135] The storage unit 1220 may also include a program / utility 12204 having a set (at least one) of program modules 12205, such program modules 12205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0136] Bus 1230 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0137] The electronic device 1200 may also communicate with one or more external devices 1240 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 1200, and / or communicate with any device that enables the electronic device 1200 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 1250. Furthermore, the electronic device 1200 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 1260. As shown, the network adapter 1260 communicates with other modules of the electronic device 1200 via a bus 1230. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 1200, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0138] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0139] Based on the same inventive concept, the embodiment of the present disclosure also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, any of the above-mentioned methods for generating product business documents based on a large model is implemented. Since the principle of solving the problem in the embodiment of the computer-readable storage medium is similar to that in the above-mentioned method embodiment, the implementation of the embodiment of the computer-readable storage medium can refer to the implementation of the above-mentioned method embodiment, and the repeated parts will not be repeated.

[0140] More specific examples of computer-readable storage media in the present disclosure may include, but are not limited to, an electrical connection having one or more conductors, 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), 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.

[0141] In the present disclosure, a computer readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, wherein a readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A readable signal medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0142] Alternatively, the program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the foregoing.

[0143] In a specific implementation, the program code for performing the operations of the present disclosure may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., using an Internet service provider to connect through the Internet).

[0144] Based on the same inventive concept, a computer program product is also provided in the embodiments of the present disclosure, including a computer program product, including: a computer program or an instruction, which, when executed by a processor, implements any one of the methods for generating product business documents based on a large model in the above method embodiments. Since the principle of solving the problem in the computer program product embodiment is similar to that in the above method embodiment, the implementation of the computer program product embodiment can refer to the implementation of the above method embodiment, and the repeated parts will not be repeated.

[0145] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0146] In addition, although the steps of the method in the present disclosure are described in a specific order in the drawings, this does not require or imply that the steps must be performed in this specific order, or that all the steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps, etc.

[0147] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0148] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the appended claims.

Claims

1. A method for generating product business documents based on a large model, characterized in that: include: Obtain product business description information of the target product; Determine, based on the product business description information, various catalog chapters included in the product business document to be generated; According to each catalog chapter included in the product business document to be generated, the pre-trained product big model is used, combined with the pre-built product business specification vector library and product knowledge graph, to generate the chapter content included in each catalog chapter, wherein the product business specification vector library includes: product business specification information of multiple stock products; the product knowledge graph includes: full product information of multiple stock products; Generate product business documents for the target product based on the chapter contents contained in each catalog chapter.

2. The method for generating product business documents based on a large model according to claim 1, characterized in that: According to each catalog chapter of the product business document to be generated, the pre-trained product model is used in combination with the pre-built product business specification vector library and product knowledge graph to generate the chapter content contained in each catalog chapter, including: According to the product business description information of the target product, searching for the product business specification information of the target product from a pre-built product business specification vector library; According to each catalog chapter contained in the product business document to be generated, a pre-trained product macro model is used to conduct multiple question-and-answer dialogues with the target user, and according to the dialogue content output by the product macro model in each question-and-answer dialogue, a product knowledge graph subgraph of the target product is dynamically constructed, wherein the product knowledge graph subgraph is used to obtain product association information of the target product; Fusing the product business specification information of the target product with the product knowledge graph subgraph to obtain fusion information of the target product; The various catalog chapters contained in the product business document to be generated, the fusion information of the target product and the preset prompt words are input into the product macro model, and the chapter contents contained in each catalog chapter in the product business document to be generated are output.

3. The method for generating product business documents based on a large model according to claim 2, characterized in that: According to the various catalog chapters contained in the product business document to be generated, a pre-trained product model is used to conduct multiple question-and-answer dialogues with the target user, and according to the dialogue content output by the product model in each question-and-answer dialogue, a product knowledge graph subgraph of the target product is dynamically constructed, including: The dialogue content output by the product macro model in each question-and-answer dialogue is used as a node to search for product-related information from the product knowledge graph and generate a corresponding product knowledge graph subgraph; According to the product knowledge graph subgraph corresponding to the dialogue content output by the product big model in the historical question-and-answer dialogue, the product knowledge graph subgraph corresponding to the dialogue content output in the current question-and-answer dialogue is updated.

4. The method for generating product business documents based on a large model according to claim 1, characterized in that: Before generating the chapter contents contained in each catalog chapter according to each catalog chapter contained in the to-be-generated product business document, using a pre-trained product macro model in combination with a pre-built product business specification vector library and a product knowledge graph, the method further includes: Get product private domain data; According to the private domain data of the product, the pre-trained large language model is fine-tuned to obtain the large model of the product.

5. The method for generating product business documents based on a large model according to claim 1, characterized in that: Before generating the chapter contents contained in each catalog chapter according to each catalog chapter contained in the to-be-generated product business document, using a pre-trained product macro model in combination with a pre-built product business specification vector library and a product knowledge graph, the method further includes: Obtain product business specification information for multiple stock products; A product business specification vector library is constructed according to the product business specification information of the multiple stock products.

6. The method for generating product business documents based on a large model according to claim 1, characterized in that: Before generating the chapter contents contained in each catalog chapter according to each catalog chapter contained in the to-be-generated product business document, using a pre-trained product macro model in combination with a pre-built product business specification vector library and a product knowledge graph, the method further includes: Get the full product information of multiple stock products; A product knowledge graph is constructed based on the full product information of the multiple stock products.

7. The method for generating product business documents based on a large model according to claim 1, characterized in that: After generating the chapter contents contained in each catalog chapter according to each catalog chapter contained in the to-be-generated product business document by using a pre-trained product macro model in combination with a pre-built product business specification vector library and a product knowledge graph, the method further includes: Show target users the contents of each directory chapter; Receiving an editing operation of the target user on the chapter contents of one or more target chapters; According to the editing operation, the chapter content under the corresponding target chapter is adjusted.

8. The method for generating product business documents based on a large model according to any one of claims 1 to 7, characterized in that: The product business document is a product business specification document.

9. A product business document generation device based on a large model, characterized in that: include: An information collection module is used to obtain product business description information of the target product; A product business document catalog chapter determination module, used to determine the catalog chapters included in the product business document to be generated according to the product business description information; The product business document intelligent writing module is used to generate the chapter contents contained in each catalog chapter according to each catalog chapter contained in the product business document to be generated, using a pre-trained product big model, combined with a pre-built product business specification vector library and a product knowledge graph, wherein the product business specification vector library contains: product business specification information of multiple stock products; the product knowledge graph contains: full product information of multiple stock products; The product business document generation module is used to generate the product business document of the target product according to the chapter contents contained in each catalog chapter.

10. An electronic device, characterized in that: include: processor; as well as A memory, configured to store executable instructions of the processor; The processor is configured to execute the method for generating a product business document based on a large model as described in any one of claims 1 to 8 by executing the executable instructions.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for generating a product business document based on a large model as described in any one of claims 1 to 8 is implemented.

12. A computer program product comprising: A computer program or instruction, characterized in that when the computer program or instruction is executed by a processor, it implements the method for generating product business documents based on a large model as described in any one of claims 1 to 8.

Images