College student competition intelligent auxiliary system based on AI large model

Abstract

The invention belongs to the technical field of education informatization, and discloses an AI large model-based college student competition intelligent auxiliary system, which provides a one-stop intelligent tutoring service for college student competition through a material uploading module, a multi-modal analysis module, a case library retrieval module, an AI project diagnosis module, a material generation module and the like. The system can automatically identify text documents, images and code contents and generate structured data of projects; and based on a preset case library, project similarity matching is realized, and reference experience information is extracted. The AI large model is used for performing multi-dimensional scoring on the project, a score deduction basis is provided, operable improvement suggestions are generated, and meanwhile, competition documents such as road performance materials, project abstracts or commercial schedules can be automatically generated. According to the method, the project preparation efficiency can be remarkably improved, the material quality is optimized, accurate and personalized competition support is provided for college students, and the method has good practical value and popularization significance.

Description

Technical Field

[0001] This invention belongs to the field of educational informatization technology, specifically relating to an intelligent auxiliary system for college student competitions based on a large AI model. Background Technology

[0002] With the booming development of various college student competitions, more and more university students are participating in competitions in fields such as creativity, innovation, and scientific research. However, although college student competitions provide students with opportunities to hone and showcase their talents, they still face a series of challenges in the actual participation process. First, student teams often lack professional guidance, especially in areas such as project selection and the writing and optimization of competition materials, often relying on the experience of mentors or team members, making the competition preparation process highly dependent on human resources. Second, most current competition tutoring systems mainly focus on specific knowledge transmission and answer simulation, lacking the ability to automatically generate, intelligently analyze, and provide real-time feedback for competition project materials.

[0003] Against this backdrop, existing tutoring tools often lack integrated functions such as competition project management, intelligent analysis, and material generation, causing students to spend a significant amount of time and energy on preparation without effectively improving project quality and participation efficiency. Furthermore, low levels of intelligence and insufficient personalization are also common problems, making it difficult to meet the specific needs of different students and different tracks.

[0004] Therefore, there is an urgent need for an intelligent assistance system based on artificial intelligence technology, especially based on large models, to provide college students with one-stop assistance services for competition projects, helping students from the topic selection stage, project writing, material optimization to final review feedback, forming a comprehensive and intelligent project guidance and support. Summary of the Invention

[0005] To overcome the above-mentioned technical problems, this invention provides an intelligent assistance system for college student competitions based on a large AI model.

[0006] The present invention adopts the following technical solution: A smart assistance system for university student competitions based on a large AI model includes: The materials upload module is used to receive competition materials uploaded by users. The materials include text documents, image files, code compressed packages, and reference materials. The multimodal parsing module is used to perform content recognition, structure extraction, and information aggregation on the material to form structured project data containing project summary, key technical points, data structure, and document hierarchy information; The case study retrieval module is used to retrieve similar projects from a preset competition project case study library based on the structured project data and output the associated results; The AI ​​project diagnosis module is used to input the structured project data into the AI ​​big model and generate project diagnosis results including multi-dimensional scores, reasons for deductions and improvement suggestions through preset prompt templates. The material generation and export module is used to automatically generate project summaries, roadshow materials, or business plans based on the project diagnostic results and export them as target format files.

[0007] Preferably, the multimodal parsing module includes: The text parsing unit is used to parse the paragraph structure, chapter titles, keywords, and abstracts of PDF and Docx documents; The image parsing unit is used to perform OCR recognition on images and identify chart types; The code parsing unit is used to identify the code language type, file structure, dependencies, and functional modules.

[0008] Preferably, the case library retrieval module includes: A vectorization unit is used to convert the structured project data into a vector representation; The similarity retrieval unit is used to perform similarity retrieval in a pre-built project vector library; The rule filtering unit is used to filter search results based on competition track requirements, project category, or user role.

[0009] Preferably, the AI ​​project diagnostic module is configured to automatically load the corresponding scoring dimensions according to the scoring system of different competitions, and output at least five scoring indicators such as innovation, feasibility, technological maturity, team capability, and market prospects.

[0010] Preferably, the material generation and export module is further configured to generate a visual report containing analysis charts, scoring radar charts, and project structure diagrams based on the project diagnostic results.

[0011] This invention also discloses an intelligent assistance method for college student competitions based on a large AI model, comprising the following steps: S1: Receive and store competition materials uploaded by users; S2: Perform multimodal analysis on the competition materials to generate structured project data; S3: Retrieve similar projects from a preset case library based on the structured project data; S4: Input the structured project data into the AI ​​big model to generate project scores, reasons for deductions, and improvement suggestions; S5: Output project analysis reports, presentation materials, or business plans based on the generated project diagnostic results.

[0012] Preferably, step S2 includes paragraph level recognition, keyword extraction and key sentence extraction for text documents, OCR recognition for image materials, and language recognition and module structure analysis for code materials.

[0013] Preferably, step S3 includes: S31: Vectorize the project summary, technical points, and target scenario; S32: Perform vector similarity retrieval in a case library containing at least 100,000 project records; S33: Filter search results based on specified competition track rules.

[0014] Preferably, step S4 includes generating a JSON-formatted structured result based on a preset prompt word template, requiring the AI ​​large model to output a score that includes at least five dimensions, corresponding deduction criteria, and actionable improvement measures.

[0015] Preferably, step S5 further includes automatically generating a PPT text framework, project summary, or visual analysis chart based on the project diagnostic results, and exporting it as competition materials in PDF or Docx format.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention's auxiliary system leverages the deep learning capabilities of large-scale artificial intelligence models, combined with multimodal data analysis technology and intelligent matching algorithms, to improve the efficiency and quality of preparation for university students' competition projects. The system not only provides intelligent guidance throughout the entire process, from project selection to material optimization, but also offers personalized suggestions and guidance tailored to the specific needs of different competition tracks.

[0017] First, this system can intelligently match the most suitable competition topics based on the student's and team's background, interests, and skills. It also incorporates data from past award-winning projects to help students quickly identify project directions that align with their abilities and the requirements of the competition. This process significantly reduces the blind spots and wasted time in the topic selection stage of traditional competitions.

[0018] Secondly, by performing multimodal analysis on the uploaded competition materials, the system can intelligently extract key content from the project documents, identify technical highlights, innovative points, and potential problems, and provide students with clear and actionable optimization suggestions. Students no longer need to rely on tedious manual modifications and repeated discussions; the system can directly provide multi-dimensional scoring and item-by-item suggestions, helping students improve project quality in a short period of time.

[0019] In addition, the system has the ability to automatically generate reports, presentation slides, and business plans. This not only saves students a lot of writing and formatting time, but also ensures the standardization and professionalism of competition materials, allowing students to focus more on the innovation and improvement of the project itself.

[0020] Most importantly, through the real-time feedback mechanism of the AI ​​large-scale model, the system can quickly evaluate projects and provide precise improvement suggestions after students submit their materials, truly achieving a "self-optimization" closed loop. In this way, students can continuously improve and enhance the quality of their projects during the competition preparation process, greatly improving the efficiency of competition preparation and the probability of winning.

[0021] In summary, this invention innovatively combines artificial intelligence with university student competitions, which not only greatly improves the efficiency and intelligence of project preparation, but also provides university students with more personalized and precise tutoring services, and has significant academic value and practical application significance. Attached Figure Description

[0022] Figure 1 This is a system architecture and workflow diagram of the present invention. Detailed Implementation

[0023] The embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings. Unless otherwise specified, the raw materials and equipment used can be purchased from the market or are commonly used in the art. The methods in the embodiments, unless otherwise specified, are conventional methods in the art. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0024] AI-based large-scale model-based intelligent assistance system for college student competitions I. Overall Architecture This embodiment provides an intelligent assistance system for university student competitions based on a large AI model. The system is deployed in a server environment according to a functional layering approach, including a front-end presentation layer, a business logic layer, an intelligent analysis layer, and a data storage layer.

[0025] The system runs on a Linux server (such as Ubuntu or CentOS) and provides HTTPS services externally through Nginx as a reverse proxy server. The front end uses a web interface to display operation pages to users; the business logic layer receives user requests and manages user and project data through the backend program; the intelligent analysis layer is responsible for file parsing, case matching, project evaluation, and material generation; and the data storage layer is used to store project information, case library, user information, and analysis results.

[0026] The core functions of the intelligent analysis layer are driven by a large AI model. By combining structured prompts with multimodal input, it can understand the content of uploaded project materials, extract key points, score and generate optimization suggestions, thereby providing users with complete competition coaching capabilities.

[0027] II. System Deployment and Operation Process After system deployment, users log in as "students," "mentors," or "reviewers" and access the system homepage to perform operations. The complete system operation process includes stages such as: material uploading, material parsing, case matching, project diagnosis, automatic material generation, and result export.

[0028] 1. Upload Materials Users can upload various types of competition-related materials, including but not limited to: Project specifications (PDF, Docx) Project Presentation PPT References, papers, patent documents Prototype code compressed package Images, flowcharts, etc. The system stores files in the server's file area and simultaneously generates a unique file identifier, which is then written to the database.

[0029] 2. Multimodal Material Analysis The intelligent analysis layer performs multimodal parsing of the file content, mainly including: Text document parsing: Automatically identifies document titles, chapter structures, paragraph content, key sentences, and keywords, and extracts summaries and project highlights.

[0030] Image analysis: OCR technology is used to recognize the text content in the image; the chart structure is identified by type, and key indicators are extracted from the chart.

[0031] Code analysis: Automatically identifies the code language, main modules, dependency files, function descriptions, code structure, and project size.

[0032] Paper / Patent Analysis: Extracting information such as abstracts, innovative points, technical solutions, experimental data, and claim elements.

[0033] After parsing, the system integrates the above information into a structured data package for subsequent AI analysis and decision-making.

[0034] 3. Matching past case studies The system contains over 100,000 outstanding project examples from university student competitions, covering multiple events and tracks. The system performs feature vectorization on the parsed structured project data and achieves the following functions through vector retrieval and rule filtering: Topic selection and track matching Learning from similar projects Comparative Analysis of Past Award-Winning Projects Strengths and weaknesses identification corresponding to the target competition evaluation criteria The matching results are presented visually, including similarity scores, case rankings, and case highlight extraction.

[0035] III. AI-Driven Project Diagnosis and Intelligent Scoring After completing the material analysis, the system inputs the structured information into the AI ​​model and, using professional prompt templates, requests the model to generate a complete project diagnostic result, including: 1. Multi-dimensional scoring Depending on the requirements of different competition tracks, the system automatically loads the corresponding scoring dimensions, such as innovation, feasibility, team capability, market prospects, and technological maturity. The AI ​​big model outputs a score of 1-10 for each item based on the uploaded materials.

[0036] 2. Reasons for deduction and chain of evidence The system requires the large model to provide reasons for deductions based on the content of the original text. For example: "Insufficient experimental data" "The technical roadmap is not clear enough" "The business model lacks cost estimation" "Prototype validation did not demonstrate adaptability to the target scenario." Furthermore, supporting evidence is provided through paragraph summaries of cited materials, thereby enhancing the reliability of the evaluation.

[0037] 3. Actionable improvement suggestions The system further outputs targeted and actionable optimization measures, such as: Size of the dataset to be supplemented Upgrade direction of core algorithms or system modules Specific guidance on modifying the project's logical framework Provide an adoptable structural example by comparing with previous case studies. Methods to improve market analysis or experimental plans These suggestions are directly used to guide users in improving their project materials.

[0038] IV. Material Generation and Result Export After completing the project diagnosis, if the user selects the material generation function, the system can generate the following content based on the intelligent analysis results: Business plan draft: includes modules such as project background, market pain points, technical principles, business model, competitive analysis, and implementation path.

[0039] Roadshow PPT Content Framework: Automatically generates a PPT text draft with titles, key points, and explanations.

[0040] Visualized reports: such as scoring radar charts, multi-dimensional comparison charts, project score trends, etc.

[0041] The system offers a one-click export function, which can generate official PDF or Docx files for users to download, and can also be used as the basis for submitting competition materials.

[0042] V. Examples of Specific Application Scenarios The following provides an application example of the system in a real-world scenario to further illustrate the technical implementation of the present invention.

[0043] Example: A team is preparing to participate in the "Challenge Cup" innovation project competition. The team uploaded project specifications, screenshots of experimental data, code prototype packages, and references.

[0044] The system automatically parses the contents of the instruction manual and extracts information such as "technical innovation points", "experimental methods", "data sources" and "expected application scenarios".

[0045] The system retrieves five previous award-winning projects with similar technical directions from the case library and generates a "Project Comparison Analysis Table".

[0046] AI-powered large-scale models score projects based on the content of the materials, for example: Innovation: 7 points Feasibility: 6 points Market prospects: 8 points Technology maturity: 5 points Team ability: 6 points The system automatically generates executable improvement plans based on the scores, including: Supplement at least three types of control experiments Clearly define the core algorithm parameter settings Supplementing potential user numbers and market forecasting Propose a more logical project implementation plan with time nodes. Users can choose to automatically generate PPT text frames and project summaries for subsequent defense preparation.

[0047] The entire process is completed automatically within the system, and users do not need to have professional document writing or project review experience.

[0048] VI. Implementation Results The system provided in this embodiment, through multimodal analysis and collaboration with large AI models, enables university student competition projects to form a closed-loop process from topic selection and diagnosis to material generation. Specific effects include: Improve project polishing efficiency Provide personalized guidance for users at different levels Supports automatic adaptation across multiple events and tracks. Improve content generation quality and consistency Reduce the workload of human tutors and increase the coverage of guidance. In summary, this solution can significantly improve the conceptual quality, documentation quality, and overall competition performance of university students' projects in real-world competition scenarios.

[0049] Although embodiments of the present invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to the above embodiments without departing from the principles and spirit of the present invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A smart auxiliary system for university student competitions based on a large AI model, characterized in that: include: The materials upload module is used to receive competition materials uploaded by users. The materials include text documents, image files, code compressed packages, and reference materials. The multimodal parsing module is used to perform content recognition, structure extraction, and information aggregation on the material to form structured project data containing project summary, key technical points, data structure, and document hierarchy information; The case study retrieval module is used to retrieve similar projects from a preset competition project case study library based on the structured project data and output the associated results; The AI ​​project diagnosis module is used to input the structured project data into the AI ​​big model and generate project diagnosis results including multi-dimensional scores, reasons for deductions and improvement suggestions through preset prompt templates. The material generation and export module is used to automatically generate project summaries, roadshow materials, or business plans based on the project diagnostic results and export them as target format files.

2. The system according to claim 1, characterized in that, The multimodal parsing module includes: The text parsing unit is used to parse the paragraph structure, chapter titles, keywords, and abstracts of PDF and Docx documents; The image parsing unit is used to perform OCR recognition on images and identify chart types; The code parsing unit is used to identify the code language type, file structure, dependencies, and functional modules.

3. The system according to claim 1, characterized in that, The case database retrieval module includes: A vectorization unit is used to convert the structured project data into a vector representation; The similarity retrieval unit is used to perform similarity retrieval in a pre-built project vector library; The rule filtering unit is used to filter search results based on competition track requirements, project category, or user role.

4. The system according to claim 1, characterized in that, The AI ​​project diagnostic module is configured to automatically load the corresponding scoring dimensions according to the scoring system of different competitions, and output at least five scoring indicators such as innovation, feasibility, technological maturity, team capability, and market prospects.

5. The system according to claim 1, characterized in that, The material generation and export module is further configured to generate a visual report based on the project diagnostic results, including analysis charts, scoring radar charts, and project structure diagrams.

6. A method for intelligent assistance in university student competitions based on a large AI model, characterized in that: Includes the following steps: S1: Receive and store competition materials uploaded by users; S2: Perform multimodal analysis on the competition materials to generate structured project data; S3: Retrieve similar projects from a preset case library based on the structured project data; S4: Input the structured project data into the AI ​​big model to generate project scores, reasons for deductions, and improvement suggestions; S5: Output project analysis reports, presentation materials, or business plans based on the generated project diagnostic results.

7. The method according to claim 6, characterized in that, Step S2 includes paragraph level recognition, keyword extraction and key sentence extraction for text documents, OCR recognition for image materials, and language recognition and module structure analysis for code materials.

8. The method according to claim 6, characterized in that, Step S3 includes: S31: Vectorize the project summary, technical points, and target scenario; S32: Perform vector similarity retrieval in a case library containing at least 100,000 project records; S33: Filter search results based on specified competition track rules.

9. The method according to claim 6, characterized in that, Step S4 includes generating a JSON-formatted structured result based on a preset prompt word template, requiring the AI ​​large model to output a score that includes at least five dimensions, corresponding deduction criteria, and actionable improvement measures.

10. The method according to claim 6, characterized in that, Step S5 further includes automatically generating a PPT text framework, project summary, or visual analysis chart based on the project diagnostic results, and exporting it as competition materials in PDF or Docx format.

Images