A method for intelligently generating revision suggestions for paper analysis based on a large model
Through the method of intelligently generating modification opinions based on large-scale paper analysis, the problem that existing technology is difficult to efficiently and automatically understand and extract key contents of academic papers is solved, and efficient paper analysis and multi-format output are achieved across fields and across languages.
Patent Information
- Application Number
- CN202411934816.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-26
- Publication Date
- 2025-04-08
AI Technical Summary
The existing technology is difficult to efficiently and automatically understand and extract key contents of academic papers, and cannot meet actual needs.
The method of intelligently generating modification opinions based on large-model paper analysis includes paper content preprocessing, language model fine-tuning, content hierarchical analysis, abstract generation, knowledge graph construction, and multilingual support and output.
It realizes automated and accurate in-depth understanding and key information extraction of academic papers, generates efficient summary reports and knowledge graphs, which are highly adaptable across fields and across languages, and can generate multi-level and multi-format analysis results based on specific needs.
Smart Images
Figure CN119782506A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of natural language processing and artificial intelligence, and specifically to a method for intelligently generating revision suggestions for paper analysis based on a large model. Background Art
[0002] At present, paper analysis mainly relies on traditional text mining technology or rule-based methods. These methods usually require the construction of complex feature extraction rule bases and perform poorly when dealing with papers of different disciplines or language styles. In addition, although recent paper content analysis tools, such as topic models and keyword extraction technology, have made breakthroughs in semantic extraction, they are insufficient in understanding the logical relationships in long texts and extracting diverse expressions.
[0003] With the development of large language models, text analysis technology based on large models has demonstrated powerful language understanding and reasoning capabilities. However, there is currently no system specifically designed to implement paper analysis and summarization. Existing solutions often require manual annotation of data sets for fine-tuning, which is a complex and insufficiently targeted process. It is impossible to efficiently and automatically understand the overall situation of the paper and extract key content, and it is difficult to meet actual needs.
[0004] Therefore, a method for intelligently generating revision suggestions for paper analysis based on a large model is designed to overcome the above problems. Summary of the invention
[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and to provide a method for intelligently generating modification suggestions for paper analysis based on a large model. The present invention can automatically and accurately perform in-depth understanding and key information extraction of academic papers, quickly generate summary reports and knowledge graphs, and provide researchers with efficient auxiliary tools. The present invention has the characteristics of strong cross-domain and cross-language adaptability, and can generate multi-level and multi-format analysis results according to specific needs.
[0006] The present invention is implemented by the following technical solution: a method for intelligently generating revision suggestions for paper analysis based on a large model, the method mainly comprising the following steps: 1. Paper content preprocessing The input paper content is divided into data types, processed differently according to the type, and then divided into chapters. At the same time, the formatting elements in the text are detected and standardized for subsequent analysis; 2. Language Model Fine-tuning Based on the open source big model, fine-tuning is done using annotated paper corpus to improve the ability to understand the unique context of academic papers; 3. Content layer analysis Input the fine-tuned macro model by chapter and paragraph, use specific prompt word templates to understand and analyze the paper in layers, and extract the core ideas and supporting arguments of each part; IV. Abstract Generation Combined with the full text analysis results, a concise paper abstract is generated based on the large model, supporting users to choose a detailed abstract or a long summary; 5. Knowledge Graph Construction Extract core entities and their relationships from the paper and build a structured knowledge graph to help users intuitively understand the main contributions and research context of the paper; 6. Multi-language support and output It supports input analysis and output generation in multiple languages, and provides analysis reports in multiple formats including PDF, PPT, and WORD.
[0007] As a preference: the data types in step one include text data, table data, and picture data.
[0008] Preferably, the formatting elements in step 1 include titles, paragraphs, tables, and formulas.
[0009] Preferably, the core entities in step five include methods, data, and conclusions.
[0010] The beneficial effects of the present invention are as follows: Compared with the prior art, the present invention has the following beneficial effects: 1. Efficiency: The deep analysis and automatic summarization functions based on the large model significantly improve the efficiency of paper reading and understanding, and researchers can grasp the content of the paper in a short time.
[0011] 2. Accuracy: Combined with the strong language understanding ability of the large model, the system can accurately extract key information from the paper and provide high-quality summaries and analysis reports.
[0012] 3. Cross-domain adaptability: The system supports the analysis of multidisciplinary papers and is applicable to a wide range of research scenarios.
[0013] 4. Versatility: In addition to content summaries, it can also generate knowledge graphs to help researchers quickly grasp the research contributions and logical relationships of the paper.
[0014] 5. Promotion prospects: The system is suitable for universities, scientific research institutions and corporate research departments, and can be widely used in digital libraries, scientific research workflow platforms and other environments. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION
[0016] In order to enable those skilled in the art to more clearly understand the objectives, technical solutions and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments.
[0017] In the description of the present invention, it is necessary to understand that the orientation or position relationship indicated by the terms such as "up", "down", "left", "right", "inside", "outside", "horizontal" and "vertical" is based on the orientation or position relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention, rather than indicating or implying that the device or original referred to must have a specific orientation, and therefore cannot be understood as a limitation on the present invention.
[0018] The present invention will be described in detail below with reference to the accompanying drawings: Figure 1 As shown, a method for intelligently generating revision suggestions for paper analysis based on a large model, the method mainly includes the following steps: 1. Paper content preprocessing The input paper content is divided into data types, processed differently according to the type, and then divided into chapters. At the same time, the formatting elements in the text (such as titles, paragraphs, tables, formulas, etc.) are detected and standardized for subsequent analysis; 2. Language Model Fine-tuning Based on the open source big model, fine-tuning is done using annotated paper corpus to improve the ability to understand the unique context of academic papers; 3. Content layer analysis Input the fine-tuned macro model by chapter and paragraph, use specific prompt word templates to understand and analyze the paper in layers, and extract the core ideas and supporting arguments of each part; IV. Abstract Generation Combined with the full text analysis results, a concise paper abstract is generated based on the large model, supporting users to choose a detailed abstract or a long summary; 5. Knowledge Graph Construction Extract core entities (such as methods, data, and conclusions) and their relationships from papers to build a structured knowledge graph, so that users can intuitively understand the main contributions and research context of the paper; 6. Multi-language support and output Supports input analysis and output generation in multiple languages, and provides analysis reports in multiple formats such as PDF, PPT, WORD, etc. The data types in step 1 include text data, table data, and image data. The formatting elements in step 1 include titles, paragraphs, tables, and formulas. The core entities in step 5 include methods, data, and conclusions.
[0019] Anyone familiar with the technology may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.
Claims
1. A method for intelligently generating revision suggestions based on paper analysis based on a large model, characterized by: The method mainly comprises the following steps:
1. Paper content preprocessing The input paper content is divided into data types, processed differently according to the type, and then divided into chapters. At the same time, the formatting elements in the text are detected and standardized for subsequent analysis; 2. Language Model Fine-tuning Based on the open source big model, fine-tuning is done using annotated paper corpus to improve the ability to understand the unique context of academic papers; 3. Content layer analysis Input the fine-tuned macro model by chapter and paragraph, use specific prompt word templates to understand and analyze the paper in layers, and extract the core ideas and supporting arguments of each part; IV. Abstract Generation Combined with the full text analysis results, a concise paper abstract is generated based on the large model, supporting users to choose a detailed abstract or a long summary; 5. Knowledge Graph Construction Extract core entities and their relationships from the paper and build a structured knowledge graph to help users intuitively understand the main contributions and research context of the paper; 6. Multi-language support and output It supports input analysis and output generation in multiple languages, and provides analysis reports in multiple formats including PDF, PPT, and WORD.
2. The method for intelligently generating revision suggestions for paper analysis based on a large model according to claim 1 is characterized by: The data types in step 1 include text data, table data, and picture data.
3. The method for intelligently generating revision suggestions for paper analysis based on a large model according to claim 1 is characterized in that: The formatting elements in step 1 include titles, paragraphs, tables, and formulas.
4. The method for intelligently generating revision suggestions based on paper analysis based on a large model according to claim 1 is characterized by: The core entities in step five include methods, data, and conclusions.