CN117494672B - Method, device and computer-readable storage medium for generating industry documents - Google Patents

Abstract

The present invention relates to the field of text processing technology, and discloses a method, device and computer-readable storage medium for generating industry documents, the method comprising: in response to a first instruction from a user, obtaining a corresponding specification document and a document generation requirement; generating a first industry document based on the specification document and the document generation requirement; in response to a second instruction from a user, extracting evaluation rules based on a preset industry database and specification document; analyzing and evaluating the first industry document based on the evaluation rules, and generating an optimization plan; in response to a confirmation instruction from a user, generating a second industry document based on the optimization plan. After generating the first industry document through the specification document and the document generation requirement, the present invention can optimize the first industry document according to the second instruction from the user, thereby obtaining a second industry document, which can improve the efficiency of searching for data and information while improving the integrity of industry documents.

Description

Method, device and computer-readable storage medium for generating industry documents

Technical Field

The present invention relates to the technical field of text processing, and in particular to a method and device for generating industry documents, and a computer-readable storage medium.

Background technique

The generation of industry documents (such as bidding documents, project acceptance documents, etc.) mainly relies on manual writing, which not only consumes a lot of time and manpower costs, but is also prone to writing errors. In addition, traditional manually written industry documents often have problems such as inconsistent styles and non-standard standards, which is not conducive to standardized management of the industry.

With the continuous development of computer science and technology, intelligent text output based on artificial intelligence has received a lot of attention. It can understand and analyze text content and automatically generate texts needed by humans, which has high application value for a large number of document processing tasks in the industry. However, the commonly used automatic generation solutions now basically describe document recognition based on optical character recognition (OCR) technology; file parsing based on natural language processing (NLP) technology to help identify bid collusion; there may be database-based rapid filling of tender documents, but the actual content still needs to be filled in manually. It can be seen that the relevant technology only improves the efficiency of searching for data and information, and cannot automatically generate complete industry documents.

Summary of the invention

In view of this, the present invention provides a method, device and computer-readable storage medium for generating industry documents to solve the technical problem in the related technology that it only improves the efficiency of searching data and information but cannot automatically generate complete industry documents.

In a first aspect, the present invention provides a method for generating an industry document, the method comprising: in response to a first instruction from a user, obtaining a corresponding specification document and document generation requirements; generating a first industry document based on the specification document and the document generation requirements; in response to a second instruction from a user, extracting evaluation rules based on a preset industry database and specification documents; analyzing and evaluating the first industry document based on the evaluation rules, and generating a first optimization plan; in response to a confirmation instruction from the user, generating a second industry document based on the first optimization plan.

The industry document generation method provided in this embodiment can optimize the first industry document according to the user's second instruction to obtain a second industry document after generating the first industry document through standard documents and document generation requirements, thereby improving the efficiency of searching for data and information while improving the integrity of industry documents.

In an optional embodiment, before generating an industry document based on the first optimization plan in response to a user's confirmation instruction, the method also includes: adjusting the evaluation rules in response to the auxiliary evaluation rules confirmed by the user; analyzing and evaluating the first industry document based on the adjusted evaluation rules, and generating a second optimization plan.

The method for generating industry documents provided in this embodiment allows users to confirm auxiliary evaluation rules and adjust evaluation rules based on the auxiliary evaluation rules, so that the evaluation rules can focus on user needs, thereby improving the winning rate of industry documents.

In an optional embodiment, a first industry document is generated based on a specification document and document generation requirements, including: parsing the specification document through a natural language processing method to extract the information to be filled in the specification document; based on the document generation requirements and a preset industry database, extracting industry content to generate content in the information to be filled in, thereby generating a first industry document.

The industry document generation method provided in this embodiment parses the standard document through a natural language processing method, and extracts data corresponding to the information to be filled in according to the document generation requirements and a preset industry database, so as to intelligently generate the first industry document, thereby improving the efficiency of the first industry document generation while ensuring the accuracy of the content in the first industry document.

In an optional embodiment, a specification document is parsed by a natural language processing method to extract information to be filled in the specification document, including: performing word segmentation on the specification document based on a natural language processing database to extract word information; using named entity recognition to extract document requirement information from the specification document; clustering the document requirement information to obtain a document requirement collection; extracting scoring content from the specification document based on scoring keywords; extracting scoring points from the specification document using sentiment analysis methods and semantic analysis methods; generating a scoring set based on the scoring content and scoring points; extracting format description statements from the specification document based on format keywords; extracting each output field in the format description statement using syntactic analysis; finding the field description of each output field through semantic matching, and extracting the field meaning explanation; generating information to be filled in based on the document requirement collection, the scoring set, and the field meaning explanation.

The method for generating industry documents provided in this embodiment can improve the efficiency and accuracy of extracting the information to be filled in the specification document by determining the document requirement collection, the scoring set and the field meaning interpretation.

In an optional implementation, extracting evaluation rules based on a preset industry database and specification documents includes: extracting corresponding evaluation parameters and evaluation weights from the specification documents based on the preset industry database; and generating evaluation rules based on the evaluation parameters and evaluation weights.

The method for generating industry documents provided in this embodiment can accurately determine whether the first industry document can meet the requirements for winning the bid by extracting evaluation parameters and evaluation weights in the specification document from a preset industry database.

In an optional implementation, the first industry document is analyzed and evaluated based on evaluation rules, and an optimization plan is generated, including: extracting content to be evaluated corresponding to the evaluation parameters from the first industry document; evaluating the content to be evaluated based on the evaluation rules and generating evaluation results; analyzing the content to be evaluated that needs to be adjusted based on the evaluation results; and generating a first optimization plan based on the content that needs to be adjusted.

The method for generating industry documents provided in this embodiment evaluates the content to be evaluated in the first industry document through evaluation rules, so as to determine the content with greater influence in the content to be evaluated, and then adjust the content so that the generated second industry document can better meet the bidding requirements.

In an optional embodiment, in response to the auxiliary evaluation rules confirmed by the user, the evaluation rules are adjusted, including: extracting evaluation parameters and evaluation weights corresponding to the auxiliary evaluation rules from the specification document based on a preset industry database; and adjusting the evaluation rules based on the evaluation parameters and evaluation weights.

The method for generating an industry document provided in this embodiment combines the auxiliary evaluation rules determined by the user and extracts the evaluation parameters and evaluation weights from a preset industry database, so that the evaluation rules meet the needs of the user, thereby making the generated second industry document meet the needs of the user.

In an optional implementation, the first industry document is analyzed and evaluated based on the adjusted evaluation rules, and a second optimization plan is generated, including: extracting the content to be evaluated corresponding to the evaluation parameters from the first industry document; evaluating the content to be evaluated based on the adjusted evaluation rules and generating evaluation results; analyzing the content to be evaluated that needs to be adjusted based on the evaluation results; and generating a second optimization plan based on the content that needs to be adjusted.

The method for generating industry documents provided in this embodiment evaluates the content to be evaluated in the first industry document through the adjusted evaluation rules, so as to determine the content with greater influence in the content to be evaluated, and then make adjustments to the content, so that the generated second industry document can better meet the bidding requirements.

In a second aspect, the present invention provides an industry document generation device, which includes: an acquisition module for acquiring corresponding specification documents and document generation requirements in response to a first instruction from a user; a first generation module for generating a first industry document based on the specification documents and document generation requirements; an extraction module for extracting evaluation rules based on a preset industry database and specification documents in response to a second instruction from a user; an analysis and generation module for analyzing and evaluating the first industry document based on the evaluation rules and generating an optimization plan; and a second generation module for generating an industry document based on the optimization plan in response to a confirmation instruction from a user.

In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, computer instructions are stored in the memory, and the processor executes the method for generating industry documents of the above-mentioned first aspect or any corresponding embodiment thereof by executing the computer instructions.

In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method for generating industry documents of the first aspect or any corresponding embodiment thereof.

BRIEF DESCRIPTION OF THE DRAWINGS

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

FIG1 is a schematic flow chart of a method for generating an industry document according to an embodiment of the present invention;

FIG2 is a schematic flow chart of another method for generating industry documents according to an embodiment of the present invention;

FIG3 is a schematic flow chart of another method for generating industry documents according to an embodiment of the present invention;

FIG4 is a schematic diagram of the architecture of a method for generating industry documents according to an embodiment of the present invention;

FIG5 is a flow chart of another method for generating industry documents according to an embodiment of the present invention;

6 is a structural block diagram of a device for generating industry documents according to an embodiment of the present invention;

FIG. 7 is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution 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 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 those skilled in the art without creative work are within the scope of protection of the present invention.

In related technologies, with the continuous development of computer science and technology, intelligent text output based on artificial intelligence has received a lot of attention. It can understand and analyze text content and automatically generate texts needed by humans, which has high application value for a large number of document processing tasks in the industry.

However, the automatic generation solutions commonly used now basically describe document recognition based on OCR technology; file parsing based on NLP technology to help identify bid collusion; there may be database-based rapid filling of bid documents, but the actual content still needs to be filled in manually. It can be seen that the relevant technology only improves the efficiency of searching for data and information, and cannot automatically generate complete industry documents.

Based on this, an embodiment of the present invention provides a method for generating industry documents. After generating a first industry document through standard documents and document generation requirements, the first industry document can be optimized according to the user's second instructions to obtain a second industry document, which can improve the efficiency of searching for data and information while improving the integrity of industry documents.

According to an embodiment of the present invention, an embodiment of a method for generating an industry document is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

In this embodiment, a method for generating an industry document is provided, which can be used for computer devices, such as computers, servers, etc. FIG. 1 is a flow chart of a method for generating an industry document according to an embodiment of the present invention. As shown in FIG. 1 , the process includes the following steps:

Step S101, in response to a first instruction from a user, obtaining a corresponding specification document and document generation requirements.

The first instruction may be an instruction generated by a user through a mouse click, keyboard selection, or touch screen. When the user issues a first instruction for an uploaded file (such as a bidding document), the computer device may respond to the first instruction issued by the user and obtain the corresponding specification document and document generation requirements from the uploaded document.

The specification document can be a framework for the content of the corresponding user uploaded file (such as bidding document). Specifically, the specification document can include the number of software copyrights, the number of qualification certificates, the number of honorary certificates, etc., and can also include the various functional modules of the software, hardware, operation and maintenance, etc., which are not limited here.

The document generation requirement may be a requirement for the corresponding user to upload a file (such as a bidding document) (such as the document generation format, document generation content), which is not specifically limited here. For example: the requirement for uploading a file may be to fill in the income data of the past three years, then the document generation requirement is to add the income data of the past three years to the specification document. Another example: the requirement for uploading a file may be to add intellectual property content, and the document generation requirement may be to add specific content in the intellectual property to the specification document. Among them, computer equipment can add the number of software copyrights, the number of qualification certificates, the number of honorary certificates, etc. under intellectual property according to the document generation requirement.

Step S102, generating a first industry document based on the specification document and document generation requirements.

The first industry document may be a document generated according to the document generation requirement and the specification document. As can be seen from the above content, after the specification document and the document generation requirement are determined, the computer device may add content to the specification document according to the document generation requirement to generate the first industry document.

It should be noted that the first industry document can be obtained by the Pangu NLP big model. Among them, the way to build the Pangu NLP big model is to build a company database, including the company's past bidding documents library, company honors and qualifications library, company project contract information library, company case library, company product library, etc. The Pangu NLP big model is built with the data in the company database as the training set. Then, the specification document and document generation requirements are used as input to automatically generate the first industry document that meets the project requirements.

Step S103, in response to the second instruction of the user, extracting evaluation rules based on a preset industry database and specification documents.

The second instruction may be an instruction generated by a user through a mouse click, keyboard selection, or touch screen. When the user issues the second instruction, the computer device may respond to the second instruction issued by the user and extract the evaluation rules of the corresponding specification document from the preset industry database. The evaluation rules may be how to evaluate the generated first industry document. The specific evaluation process is described in detail below.

The preset industry database may include basic information of the company, product information, general solution information of the industry to which the company belongs, etc., which are not specifically limited here. For example: the name of company A, product A produced, the name of company B, product B produced, etc.

Step S104: Analyze and evaluate the first industry document based on the evaluation rules, and generate a first optimization solution.

After extracting the evaluation rules, the computer device can analyze and evaluate the first industry document according to the evaluation rules to generate a first optimization plan. The first optimization plan can be an optimization plan for adjusting the content in the first industry document that does not meet the requirements. For example, taking the bidding document as an example, the first industry document can be a bidding document. The score of the first industry document after evaluation is 80 points, which does not meet the requirements. Therefore, it is necessary to analyze the content that does not meet the requirements according to the first industry document and provide a first optimization plan for adjusting the content.

Step S105 , in response to a confirmation instruction from the user, generating a second industry document based on the first optimization solution.

The confirmation instruction may be an instruction generated by a user through a mouse click, keyboard selection, or touch screen. When the user issues a confirmation instruction, the computer device may respond to the confirmation instruction issued by the user, adjust the first industry document through the first optimization scheme, and generate a second industry document. For example, taking the bidding document as an example, the preliminary score of the bidding document is 80 points. After comparison with the preset industry database, the first industry document price is too high for the previous bidding strategy, and further optimization is recommended.

The industry document generation method provided in this embodiment can optimize the first industry document according to the user's second instruction to obtain a second industry document after generating the first industry document through standard documents and document generation requirements, thereby improving the efficiency of searching for data and information while improving the integrity of industry documents.

In an optional implementation, in order to make the second industry document more satisfying to the needs of the user, before step S105, the method further includes:

Step a1: adjusting the evaluation rules in response to the auxiliary evaluation rules confirmed by the user.

The auxiliary evaluation rules may be rules entered by the user through mouse clicks, keyboard selections, voice or touch screen. When the user issues an auxiliary evaluation rule, the computer device may respond to the auxiliary evaluation rule issued by the user and adjust the evaluation rule through the auxiliary evaluation rule. The auxiliary evaluation rule may be a personalized suggestion for the user. For example, taking a bidding document as an example, the user wants the bidding document to win the bid for the project with a lower cost, a suitable quotation, the highest profit, etc., then the user's auxiliary evaluation rule may be sent to the computer device so that the computer device adjusts the evaluation rule.

Specifically, the above step a1 includes:

Step a11, extracting evaluation parameters and evaluation weights corresponding to the auxiliary evaluation rules from the specification document based on a preset industry database.

The auxiliary evaluation rules contain user-specific focuses, such as lower cost, highest profit, etc., which are not specifically limited here. Therefore, based on the user's auxiliary evaluation rules and the preset industry database, the evaluation parameters and evaluation weights corresponding to the auxiliary evaluation rules can be extracted from the specification document.

Step a12, adjusting the evaluation rules based on the evaluation parameters and evaluation weights.

After obtaining the evaluation parameters and evaluation weights corresponding to the auxiliary evaluation rules, the computer device needs to further adjust the evaluation rules according to the evaluation parameters and evaluation weights, that is, to modify the evaluation weights and evaluation parameters of other parts of the evaluation rules.

Step a2: analyzing and evaluating the first industry document based on the adjusted evaluation rules, and generating a second optimization solution.

The computer equipment re-analyzes and evaluates the first industry document according to the adjusted evaluation rules, that is, automatically identifies the content corresponding to the first industry document and the adjusted evaluation rules, thereby determining the content with greater impact in the first industry document. For example: taking the bidding document as an example, the first industry document can be a bidding document. Among them, after evaluating the first industry document, it is recommended to optimize the parameters for displaying the legal person relationship map. The second optimization plan can be to modify the function of displaying the parameters of the legal person relationship map implemented by the knowledge graph technology to the function of displaying the parameters of the legal person relationship map using a data visualization display method; thereby, the technical difficulty can be reduced, thereby reducing the development and operation and maintenance costs. It is recommended to reduce the quotation by 20,000 yuan, and the profit estimate will still increase by 10%, and the total score of the bid will be increased to 85 points.

Specifically, the above step a2 includes:

Step a21, extracting the content to be evaluated corresponding to the evaluation parameter from the first industry document.

Step a22, based on the adjusted evaluation rules, evaluate the content to be evaluated and generate an evaluation result.

The computer device can extract the content to be evaluated from the first industry document, and re-evaluate the content to be evaluated according to the adjusted evaluation rules to generate an evaluation result. Since the evaluation weight and evaluation parameters in the adjusted evaluation rules have changed, the generated evaluation result will change.

Step a23, analyzing the content to be evaluated that needs to be adjusted based on the evaluation results.

Step a24, generating a second optimization solution based on the content that needs to be adjusted.

According to the above evaluation results, it can be analyzed which contents in the content to be evaluated do not meet the requirements, and it is necessary to adjust the contents that do not meet the requirements, and then generate a second optimization plan based on the adjusted contents. For example: taking the bidding document as an example, the first industry document can be the bidding document. Among them, after the first industry document is evaluated, it is recommended to optimize the display of the number of software copyrights. The second optimization plan can be to modify the function of displaying the number of software copyrights by using knowledge graph technology to use data visualization to display the number of software copyrights; thereby reducing the technical difficulty, thereby reducing the development and operation and maintenance costs, and it is recommended to reduce the quotation by 20,000 yuan, and the profit estimate will still increase by 20%, and the total score of the bid will be increased to 90 points.

The method for generating industry documents provided in this embodiment allows users to confirm auxiliary evaluation rules and adjust evaluation rules based on the auxiliary evaluation rules, so that the evaluation rules can focus on user needs, thereby improving the winning rate of industry documents.

In this embodiment, a method for generating an industry document is provided, which can be used for the above-mentioned computer equipment, such as a computer, a server, etc. FIG. 2 is a flow chart of the method for generating an industry document according to an embodiment of the present invention. As shown in FIG. 2 , the process includes the following steps:

Step S201, in response to the first instruction of the user, obtain the corresponding specification document and document generation requirements. Please refer to step S101 of the embodiment shown in FIG1 for details, which will not be repeated here.

Step S202: Generate a first industry document based on the specification document and document generation requirements.

Specifically, the above step S202 includes:

Step S2021, parse the specification document through natural language processing method, and extract the information to be filled in the specification document.

Natural language processing methods work through machine learning. Machine learning systems store words and how they are combined, just like any other form of data. Phrases, sentences, and sometimes even entire books are fed into a machine learning engine, where they are processed using grammatical rules or people's real-world language habits, or both. The computer then uses this data to find patterns and infer what will happen next. Take translation software as an example: in French, "I'm going to the park" is "Je vais au parc", so machine learning predicts that "I'm going to the store" will also start with "Je vais au". Therefore, the information to be filled in the specification document can be extracted through natural language processing methods.

Specifically, the above step S2021 includes:

Step b1, perform word segmentation processing on the standard document based on the natural language processing database to extract word information.

A natural language processing database refers to a database used to store and query language-related data. Specifically, it may include the HanLP database, Jieba database, etc., which are not specifically limited here. Among them, the natural language processing database can perform word segmentation on the standard document, extract word information, and clean up the text to remove irrelevant content, such as punctuation, special symbols, etc., which are not specifically limited here.

Step b2, using named entity recognition, extracts document requirement information from the specification document.

Document requirement information can be sentences containing keywords such as need and requirement. Specifically, a rule template for requirement description can be constructed, and sentences containing document requirement information can be recognized using named entity recognition. Information such as products, functions, and objects can be extracted from the specification document, and dependency syntax analysis can be used to extract requirements expressed by the subject-object structure. Dependency syntax is used to describe the dependency relationship between words.

Step b3: cluster the document demand information to obtain a document demand collection.

The document requirement collection may be a combination of sentences containing keywords such as need and requirement. Specifically, the document requirement collection is obtained by analyzing sentences with similar semantics and performing clustering processing. Optionally, the clustering method may be a K-MEANS clustering algorithm, a mean shift clustering algorithm, a DBSCAN clustering algorithm, etc., which are not specifically limited here.

Step b4, extracting the scoring content from the specification document based on the scoring keywords.

The scoring keywords may include scoring keywords such as important, necessary, etc. Specifically, a rule template for scoring keywords is constructed, and syntactic analysis is performed on clauses introduced by words such as score, value, weight, etc. to extract scoring content.

Step b5, using sentiment analysis methods and semantic analysis methods to extract scoring points from the specification document.

Sentiment analysis methods are the process of analyzing, processing and extracting emotionally subjective texts using natural language processing and text mining techniques. Specifically, they can be naive Bayes sentiment analysis, deep learning LSTM, pre-trained rule-based VADER models, etc., which are not specifically limited here.

Semantic analysis methods use various machine learning algorithms to learn and understand the semantic content represented by a piece of text.

Computer equipment can use sentiment analysis methods to determine the scoring tendency contained in a sentence, and then use semantic analysis methods to extract scoring points from the standard document.

Step b6, generating a rating set based on the rating content and rating points.

After acquiring the scoring content and the scoring key points, the computer device may merge the scoring content and the scoring key points to generate a scoring set.

Step b7, extracting format description sentences from the specification document based on format keywords.

The format keywords may include format keywords such as input, print, etc., which are not specifically limited here. Specifically, the format description sentence is identified according to the format keyword, and the format description sentence is extracted from the specification document.

Step b8, extracting each output field in the format description statement by using syntax analysis.

Syntactic analysis can be sentence component analysis, hierarchical analysis, variation analysis, etc., which are not specifically limited here. Among them, taking sentence component analysis as an example, the format description sentence is analyzed by sentence component analysis. During the analysis, it is required to find the two central words of the entire structure at one go, a noun central word and a verb central word, as the main components of the structure, the subject and the predicate, and let the other components be attached to the subject and the predicate respectively. The analysis process is: first see the main components of the entire structure, which is the subject and which is the predicate; then see whether the verb of the predicate is a transitive verb to determine whether there is a joint component object after the predicate; finally, point out all the additional components attached before the subject and object, and before and after the predicate, and then extract each output field.

Step b9, find the field description of each output field through semantic matching, and extract the field meaning explanation.

After each output field is determined, the field description of each output field is found through semantic matching, and the field meaning explanation is extracted.

Step b10, generating information to be filled in based on the document requirement collection, the score collection and the field meaning explanation.

Since the information to be filled in is composed of a document requirement collection, a score collection and field meanings, the computer device can generate the information to be filled in after obtaining the document requirement collection, the score collection and the field meaning interpretation.

Step S2022, based on the document generation requirements and the preset industry database, extract the industry content to generate the content in the information to be filled in, and generate the first industry document.

From the above content, it can be seen that the document generation requirements can be the requirements in the uploaded file (such as document generation format, document generation content), and the preset industry database can include the company's basic information, product information, and general solution information of the industry to which the company belongs, etc. Then, according to the document generation requirements and the preset industry database, the industry content of the corresponding uploaded file can be extracted, so that the industry content can be used as the content in the information to be filled in, and the first industry document can be generated according to the content in the information to be filled in.

Step S203, in response to the second instruction of the user, extracting evaluation rules based on the preset industry database and standard documents. Please refer to step S103 of the embodiment shown in FIG. 1 for details, and will not be repeated here.

Step S204: Analyze and evaluate the first industry document based on the evaluation rule, and generate a first optimization solution. Please refer to step S104 of the embodiment shown in FIG1 for details, which will not be described here.

Step S205: In response to the user's confirmation instruction, generate a second industry document based on the first optimization solution. Please refer to step S105 of the embodiment shown in FIG1 for details, which will not be repeated here.

The industry document generation method provided in this embodiment parses the standard document through a natural language processing method, and extracts data corresponding to the information to be filled in according to the document generation requirements and a preset industry database, so as to intelligently generate the first industry document, thereby improving the efficiency of the first industry document generation while ensuring the accuracy of the content in the first industry document.

In this embodiment, a method for generating an industry document is provided, which can be used for the above-mentioned computer equipment, such as a computer, a server, etc. FIG. 3 is a flow chart of the method for generating an industry document according to an embodiment of the present invention. As shown in FIG. 3 , the process includes the following steps:

Step S301, in response to the first instruction of the user, obtain the corresponding specification document and document generation requirements. Please refer to step S201 of the embodiment shown in FIG2 for details, which will not be repeated here.

Step S302: Generate a first industry document based on the specification document and the document generation requirement. Please refer to step S202 of the embodiment shown in FIG2 for details, which will not be described in detail here.

Step S303, in response to the second instruction of the user, extracting evaluation rules based on a preset industry database and specification documents.

Specifically, the above step S303 includes:

Step S3031, extracting corresponding evaluation parameters and evaluation weights from the specification document based on a preset industry database.

The specification document includes evaluation parameters and evaluation weights corresponding to different factors. The evaluation parameters and evaluation weights corresponding to each factor can be extracted from the specification document according to the preset industry database.

Step S3032: Generate evaluation rules based on evaluation parameters and evaluation weights.

Scoring rules: F1, F2, F3, and F4 correspond to technical parameters, business parameters, bidding quotation parameters, and priority procurement policy parameters respectively; A1, A2, A3, and A4 correspond to the weights of technical parameters, business parameters, and bidding quotation parameters respectively. Total coefficient = A1×F1+A2×F2+A3×F3+A4×F4.

F1 contains sub-parameters: F11, F12, and F1n, which correspond to the technical requirements clearly required in the above-mentioned uploaded documents (such as bidding documents), such as the development and maintenance of the administrative division code change module, the upgrade and maintenance of the full population data query interface, and the terminal data encryption and decryption functions;

F2 contains sub-parameters: F21, F22, F23, F24, and F2n, which correspond to the business requirements clearly required in the above uploaded documents (such as bidding documents), such as technical capabilities, service capabilities, similar performance, etc. Among them, the technical capability parameter can be further subdivided into parameters such as F211, F212, and f21n, which respectively represent the number of software copyrights, the number of qualification certificates, the number of honorary certificates, etc.;

F3 contains sub-parameters: F31, F32, F33, F34, and F3n, which correspond to software functional modules, hardware, operation and maintenance, etc.

F4 contains sub-parameters: F41, F42, F43, F44, and F4n, which respectively correspond to the priority procurement bonus requirements clearly required in the bidding documents, such as small and medium-sized enterprises and two-type products; among them, the F1, F2, F3, and F4 coefficients are the sum of the coefficients of each sub-item.

Specifically, the above step S3032 may include:

Step c1, extracting the content to be evaluated corresponding to the evaluation parameter from the first industry document.

Step c2: Based on the evaluation rules, the content to be evaluated is evaluated to generate an evaluation result.

The computer device may extract the content to be evaluated from the first industry document. The computer device re-evaluates the content to be evaluated according to the adjusted evaluation rules to generate an evaluation result. Since the evaluation weight and evaluation parameters in the adjusted evaluation rules have changed, the generated evaluation result may change.

Step c3: analyzing the content to be evaluated that needs to be adjusted based on the evaluation results.

Step c4: generating a first optimization solution based on the content that needs to be adjusted.

According to the above evaluation results, it can be analyzed which contents in the content to be evaluated do not meet the requirements, and it is necessary to adjust the contents that do not meet the requirements, and then generate the first optimization plan based on the adjusted contents. For example: taking the bidding document as an example, the first industry document can be the bidding document. Among them, after the first industry document is evaluated, it is recommended to optimize the display of the number of software copyrights. The first optimization plan can be to modify the function of displaying the number of software copyrights by using knowledge graph technology to use data visualization to display the number of software copyrights; thereby reducing the technical difficulty, thereby reducing the development and operation and maintenance costs, and it is recommended to reduce the quotation by 20,000 yuan, the profit estimate will still increase by 20%, and the total score of the bid will be increased to 90 points.

Step S304: Analyze and evaluate the first industry document based on the evaluation rule, and generate a first optimization solution. Please refer to step S203 of the embodiment shown in FIG2 for details, which will not be described here.

Step S305: In response to the user's confirmation instruction, generate a second industry document based on the first optimization solution. Please refer to step S205 of the embodiment shown in FIG2 for details, which will not be repeated here.

The method for generating industry documents provided in this embodiment evaluates the content to be evaluated in the first industry document through evaluation rules, so as to determine the content with greater influence in the content to be evaluated, and then adjust the content so that the generated second industry document can better meet the bidding requirements.

The large model (such as the Pangu NLP large model mentioned above) adopts a multi-model collaboration solution to simulate the writing process of the first industry document and the second industry document, and provides automatic optimization functions based on the output content. In an optional implementation, FIG4 shows an architectural diagram of an industry document generation method, and FIG5 shows a process diagram of an industry document generation method.

The overall architecture is divided into: collaboration layer and model layer.

The role of the collaboration layer is to assist the model layer in completing tasks, and it is also the underlying dependency of the model layer. Specifically, the collaboration layer assigns roles to each large model. Specifically, it can define a set of roles, such as technicians, auditors, etc., and define some attributes for each role, such as role name, goals, constraints, etc., which are not specifically limited here. The context of each large language model is stored in the database for other large language models to call.

The role of the model layer is to receive user information and output the final industry document. Specifically, multiple large models exchange information through the collaboration layer, and ultimately the entire industry document output process can be completed based on user needs. Specifically, the model layer can include model A and model B. Model A: The role of a technician, responsible for outputting the first industry document according to the specification document and document generation requirements. Model B: The role of an auditor, responsible for understanding the scoring rules and scoring the output of model A according to the scoring rules. If it is below the scoring standard, a modification instruction is generated, and then model A is iterated multiple times based on the instruction and user modification suggestions until the scoring standard is met.

As shown in FIG. 4 , the file in FIG. 4 may be a bidding document. The bidding document is processed by a natural language processing method (i.e., the NLP technology in FIG. 4 ) to obtain bidding requirements, scoring criteria, and bidding formats. Among them, the bidding requirements may be document generation requirements, the bidding format may be a specification document, and the scoring criteria may be an evaluation rule. Then, the corresponding data is extracted from the company's bidding database and given to the collaboration layer, and then the collaboration layer gives the data to model A. Model A generates a draft (i.e., the first industry document) according to the bidding requirements and the bidding format, and then the first industry document is sent to model B through the collaboration layer. Model B evaluates the first industry document and gives feedback. Among them, the user can add auxiliary scoring rules, i.e., I want a "XX" bid in FIG. 4, for example: I want a bid with lower cost, appropriate quotation, and highest profit. Then, through the auxiliary scoring rules, feedback is given again, and the content of the feedback may be to modify the "XX" part, for example: modify the knowledge graph technology to a data visualization display method.

It should be noted that the collaboration layer may include conversation context and document sharing. The conversation context may be stored in a conversation database. Document sharing is used to implement document sharing between model A and model B.

As shown in FIG. 5 , the bidding document in FIG. 5 can be a first industry document and a second industry document, the bidding document can be the above-mentioned file uploaded to the computer server, the bidding requirements can be file generation requirements, and the bidding format can be a standard document. Among them, the user inputs the bidding document into the computer server, and the computer server analyzes the bidding requirements and bidding format of the bidding document through the collaboration layer to generate the first industry document, and then evaluates the first industry document through model B and gives improvement suggestions, that is, the first optimization solution. Among them, auxiliary scoring rules can be added, and the scoring rules can be modified according to the auxiliary scoring rules, and the first industry document can be re-evaluated according to the modified scoring rules, and specific optimization suggestions can be given, that is, the second optimization solution.

In this embodiment, an industry document generation device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and/or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

This embodiment provides an industry document generation device, as shown in Figure 6, including: an acquisition module 601, used to respond to a first instruction from a user to obtain a corresponding specification document and document generation requirements; a first generation module 602, used to generate a first industry document based on the specification document and the document generation requirements; an extraction module 603, used to respond to a second instruction from a user to extract evaluation rules based on a preset industry database and specification documents; an analysis generation module 604, used to analyze and evaluate the first industry document based on the evaluation rules and generate an optimization plan; a second generation module 605, used to respond to a confirmation instruction from the user to generate a second industry document based on the optimization plan.

In an optional embodiment, the above-mentioned device also includes: an adjustment module, which is used to adjust the evaluation rules in response to the auxiliary evaluation rules confirmed by the user; an evaluation generation module, which is used to analyze and evaluate the first industry document based on the adjusted evaluation rules and generate a second optimization plan.

In an optional embodiment, the first generation module 602 includes: a first extraction unit, which is used to parse the specification document through a natural language processing method to extract the information to be filled in the specification document; a second extraction unit, which is used to extract industry content to generate the content in the information to be filled in based on the document generation requirements and a preset industry database, and generate a first industry document.

In an optional embodiment, the first extraction unit includes: a first extraction subunit, which is used to perform word segmentation processing on the standard document based on a natural language processing database to extract word information; a second extraction subunit, which is used to use named entity recognition to extract document requirement information from the standard document; a clustering subunit, which is used to cluster the document requirement information to obtain a document requirement collection; a third extraction subunit, which is used to extract scoring content from the standard document based on scoring keywords; a fourth extraction subunit, which is used to extract scoring points from the standard document using sentiment analysis methods and semantic analysis methods; a first generation subunit, which is used to generate a scoring set based on scoring content and scoring points; a fifth extraction subunit, which is used to extract format description statements from the standard document based on format keywords; a sixth extraction subunit, which is used to extract each output field in the format description statement using syntactic analysis; a seventh extraction subunit, which is used to find the field description of each output field through semantic matching and extract the field meaning explanation; and a second generation subunit, which is used to generate information to be filled in based on the document requirement collection, the scoring set and the field meaning explanation.

In an optional embodiment, the extraction module 603 includes: a third extraction unit, used to extract corresponding evaluation parameters and evaluation weights from the specification document based on a preset industry database; and a first generation unit, used to generate evaluation rules based on the evaluation parameters and evaluation weights.

In an optional embodiment, the analysis generation module 604 includes: a fourth extraction unit, used to extract the content to be evaluated corresponding to the evaluation parameters from the first industry document; a second generation unit, used to evaluate the content to be evaluated based on the evaluation rules and generate an evaluation result; a first analysis unit, used to analyze the content to be evaluated that needs to be adjusted based on the evaluation result; and a third generation unit, used to generate a first optimization plan based on the content that needs to be adjusted.

In an optional embodiment, the adjustment module includes: a fifth extraction unit, used to extract evaluation parameters and evaluation weights corresponding to the auxiliary evaluation rules from the specification document based on a preset industry database; and an adjustment unit, used to adjust the evaluation rules based on the evaluation parameters and evaluation weights.

In an optional embodiment, the evaluation generation module includes: a sixth extraction unit, used to extract the content to be evaluated corresponding to the evaluation parameters from the first industry document; an evaluation unit, used to evaluate the content to be evaluated based on the adjusted evaluation rules and generate an evaluation result; a second analysis unit, used to analyze the content to be evaluated that needs to be adjusted based on the evaluation result; and a fourth generation unit, used to generate a second optimization plan based on the content that needs to be adjusted.

The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

The industry document generation device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and/or other devices that can provide the above functions.

An embodiment of the present invention further provides a computer device having the device for generating the industry document shown in FIG. 6 above.

Please refer to FIG. 7, which is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention. As shown in FIG. 7, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses for communication, and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed in the computer device, including instructions stored in or on the memory to display graphical information of the GUI on an external input/output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and/or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). In FIG. 7, a processor 10 is taken as an example.

The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

The memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiment.

The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims (8)

1. A method for generating an industry document, the method comprising:

responding to a first instruction of a user, and acquiring a corresponding specification document and a document generation requirement;

generating a first industry document based on the specification document and the document generation requirement;

Responding to a second instruction of a user, and extracting an evaluation rule based on a preset industry database and the specification document;

Analyzing and evaluating the first business document based on the evaluation rule, and generating a first optimization scheme;

Generating a second industry document based on the first optimization scheme in response to a confirmation instruction of a user;

The generating a first industry document based on the specification document and the document generation requirement comprises:

Analyzing the standard document by a natural language processing method, and extracting information to be filled in the standard document;

Extracting industry content to generate content in the to-be-filled information based on the document generation requirement and a preset industry database, and generating the first industry document;

analyzing the standard document by a natural language processing method, and extracting the information to be filled in the standard document comprises the following steps:

word segmentation processing is carried out on the standard document based on a natural language processing database, and word information is extracted;

extracting document requirement information from the canonical document by using named entity recognition;

Clustering the document demand information to obtain a document demand collection;

Extracting scoring content from the canonical document based on the scoring keywords;

extracting scoring points from the canonical document by using an emotion analysis method and a semantic analysis method;

generating a scoring set based on the scoring content and scoring points;

Extracting a format description sentence from the specification document based on a format keyword;

Extracting each output field in the format description statement by utilizing syntactic analysis;

finding out the field description of each output field through semantic matching, and extracting field meaning explanation;

and generating the to-be-filled information based on the document demand collection, the grading collection and the field meaning interpretation.

2. The method of generating an industry document according to claim 1, wherein before generating a second industry document based on the first optimization scheme in response to a confirmation instruction of a user, the method further comprises:

adjusting the evaluation rule in response to the user-confirmed auxiliary evaluation rule;

And analyzing and evaluating the first industrial document based on the adjusted evaluation rule, and generating a second optimization scheme.

3. The method for generating an industry document according to claim 1, wherein the extracting evaluation rules based on the preset industry database and the specification document comprises:

extracting corresponding evaluation parameters and evaluation weights from the specification document based on a preset industry database;

the evaluation rule is generated based on the evaluation parameters and the evaluation weights.

4. The method for generating an industry document according to claim 3, wherein the analyzing and evaluating the first industry document based on the evaluation rule and generating a first optimization scheme includes:

extracting to-be-evaluated content corresponding to the evaluation parameters from the first industry document;

Evaluating the content to be evaluated based on the evaluation rule to generate an evaluation result;

analyzing the content to be regulated in the content to be evaluated based on the evaluation result;

And generating the first optimization scheme based on the content needing to be adjusted.

5. The method of claim 2, wherein adjusting the evaluation rule in response to the user-validated auxiliary evaluation rule comprises:

Extracting evaluation parameters and evaluation weights corresponding to the auxiliary evaluation rules from the specification document based on a preset industry database;

and adjusting the evaluation rule based on the evaluation parameter and the evaluation weight.

6. The method for generating an industry document according to claim 5, wherein the analyzing and evaluating the first industry document based on the adjusted evaluation rule, and generating a second optimization scheme, comprises:

extracting to-be-evaluated content corresponding to the evaluation parameters from the first industry document;

evaluating the content to be evaluated based on the adjusted evaluation rule to generate an evaluation result;

analyzing the content to be regulated in the content to be evaluated based on the evaluation result;

And generating the second optimization scheme based on the content needing to be adjusted.

7. An apparatus for generating an industry document, the apparatus comprising:

The acquisition module is used for responding to the first indication of the user and acquiring corresponding specification documents and document generation requirements;

The first generation module is used for generating a first industry document based on the specification document and the document generation requirement;

the extraction module is used for responding to a second instruction of the user and extracting an evaluation rule based on a preset industry database and the specification document;

The analysis generation module is used for carrying out analysis and evaluation on the first business document based on the evaluation rule and generating an optimization scheme;

The second generation module is used for responding to the confirmation indication of the user and generating a second industry document based on the optimization scheme;

The first generation module includes: the first extraction unit is used for analyzing the standard document by a natural language processing method and extracting information to be filled in the standard document; the second extraction unit is used for extracting the content in the industry content generation information to be filled based on the document generation requirement and a preset industry database to generate a first industry document;

the first extraction unit includes: the first extraction subunit is used for carrying out word segmentation processing on the standard document based on the natural language processing database and extracting word information; the second extraction subunit is used for extracting document requirement information from the standard document by using named entity identification; the clustering subunit is used for carrying out clustering processing on the document demand information to obtain a document demand collection; a third extraction subunit, configured to extract scoring content from the canonical document based on the scoring keyword; a fourth extraction subunit, configured to extract a scoring gist from the canonical document using an emotion analysis method and a semantic analysis method; the first generation subunit is used for generating a scoring set based on scoring content and scoring points; a fifth extraction subunit for extracting the format description sentence from the canonical document based on the format keyword; a sixth extraction subunit for extracting each output field in the format description sentence by using syntactic analysis; a seventh extraction subunit, configured to find a field description of each output field through semantic matching, and extract a field meaning explanation; and the second generation subunit is used for generating the to-be-filled information based on the document requirement set, the grading set and the field meaning interpretation.

8. A computer-readable storage medium having stored thereon computer instructions for causing a computer to perform the method of generating an industry document according to any one of claims 1 to 6.