A complete academic paper report generation system, including backend API services and frontend web application, supporting AI-based automatic academic paper report generation.
This project aims to help researchers quickly generate high-quality academic papers or reports by integrating multiple academic literature retrieval sources and large language models (LLMs), combined with Retrieval-Augmented Generation (RAG) technology, to realize the full-process automation from literature collection to report writing.
- Advanced Academic Literature Search: Integrates multiple authoritative academic platforms such as Crossref, OpenAlex, ArXiv, and Semantic Scholar, supporting concurrent retrieval and intelligent filtering
- Retrieval-Augmented Generation (RAG): Integrates academic literature retrieval results with local knowledge base, significantly improving the accuracy and relevance of report content
- Intelligent Report Generation: LLM-based report content generation and organization, supporting multiple languages
- Local Knowledge Base Management: Supports loading, automatic segmentation, and vector indexing of PDF, DOCX and other format documents
- Custom Template Functionality: Supports users to create and use custom report templates, flexibly controlling report structure and content
- User-Friendly Web Interface: Intuitive report generation and management interface
- Docker Containerized Deployment: Simplifies deployment and environment configuration
- Python 3.9+: Main development language
- Flask 2.0+: Web framework, providing RESTful API
- Transformers: Local Embedding model, supporting document vectorization
- LangChain: RAG technology implementation, integrating retrieval results with knowledge base
- OpenRouter API: Large language model service, supporting intelligent content generation
- Multi-threaded Concurrency: Implements efficient parallel retrieval of academic literature
- Vector Database: Supports efficient storage and retrieval of local knowledge base
- Vue.js 3: Modern frontend framework
- Vite: Fast build tool
- Tailwind CSS: Responsive UI design
- Docker: Containerized deployment
- Docker Compose: Multi-service coordination management
academic_paper_generation/
├── backend/ # Backend API service
│ ├── app/ # Application core code
│ │ ├── api/ # API routes
│ │ ├── services/ # Business logic services
│ │ └── app.py # Flask application main file
│ ├── config/ # Configuration files
│ ├── database/ # Database files
│ ├── requirements.txt # Python dependencies
│ └── README.md # Backend service documentation
├── frontend/ # Frontend web application
│ ├── src/ # Source code
│ ├── public/ # Static resources
│ ├── package.json # Frontend dependencies
│ └── vite.config.js # Vite configuration
├── docker-compose.yml # Docker Compose configuration
├── Dockerfile.backend # Backend Dockerfile
├── Dockerfile.frontend # Frontend Dockerfile
└── README.md # Project documentation
- Clone the repository
git clone https://github.com/S0ra-ai/academic_paper_generation.git
cd academic_paper_generation- Configure environment variables
Copy and edit the backend environment variable file:
cp backend/.env.example backend/.env
# Edit backend/.env file, fill in API keys and other configurations- Start services
docker-compose up -dServices will start at the following addresses:
- Frontend application: http://localhost:3000
- Backend API: http://localhost:5000
cd backend
pip install -r requirements.txt
# Configure .env file
python app/app.pycd frontend
npm install
npm run dev- Intelligent Literature Retrieval: Input research topic, the system automatically retrieves relevant literature from multiple academic platforms
- Knowledge Integration: Deeply integrates academic literature retrieval results with local knowledge base content
- Structured Reports: Generates logically clear, content-rich structured reports based on integrated knowledge
- Multi-language Support: Supports report generation in multiple languages including Chinese and English
- Content Accuracy: Significantly improves the accuracy and academic relevance of generated content through RAG technology
- Multi-source Integration: Simultaneously integrates authoritative academic platforms such as Crossref, OpenAlex, ArXiv, and Semantic Scholar
- Concurrent Retrieval: Multi-threaded parallel retrieval, greatly improving search efficiency
- Intelligent Processing: Automatic filtering, sorting, and deduplication of retrieval results
- Comprehensive Coverage: Ensures access to the most relevant and latest academic research results
- Standardized Processing: Unifies literature formats from different platforms for subsequent analysis and use
- Multi-format Support: Compatible with multiple document formats such as PDF and DOCX
- Automatic Processing: Intelligent document segmentation, text extraction, and vector indexing
- Efficient Retrieval: Semantic-based fast document content retrieval
- Continuous Optimization: Supports dynamic updating and expansion of knowledge base
- Secure Storage: Local storage ensures data privacy and security
- History Records: View and manage all historically generated reports
- Content Preview: Online preview of report content
- Convenient Download: Support for report export and saving
- Template Creation: Users can create personalized report templates according to their needs
- Template Management: Support for saving, editing, and deleting templates
- Flexible Configuration: Customizable report structure, chapter layout, and content requirements
- Template Application: Select to use custom templates during report generation
- Structure Control: Precisely control the logical structure and content organization of generated reports through templates
- Reusability: Created templates can be reused, improving the consistency and efficiency of report generation
Backend API service provides the following main endpoints:
- GET /health - Check API service status
- POST /api/generate-report - Generate academic report
- Request body:
{"query": "Research topic", "limit": 6, "language": "zh"} - Response: Generated report content
- Request body:
- GET /api/reports - Get report list
- GET /api/reports/{id} - Get report details
- DELETE /api/reports/{id} - Delete report
- GET /api/get-templates - Get template list
- POST /api/upload-template - Upload new template
Detailed API documentation can be found in Backend Service Documentation.
Main configuration items (located in backend/.env):
OPENROUTER_API_KEY- OpenRouter API keySERPAPI_KEY- SerpAPI key (for certain retrieval functions)ACADEMIC_EMAIL- Academic email (for API request identification)LLM_MODEL_NAME- LLM model name to useKNOWLEDGE_DIR- Local knowledge base directorySAVE_DIR- Report save directory
Frontend configuration is mainly managed through environment variables and configuration files:
VITE_API_BASE_URL- Backend API base URL
The project provides complete Docker configuration, supporting one-click deployment:
# Build and start all services
docker-compose up -d --build
# Build and start backend service separately
docker-compose up -d --build backend
# Build and start frontend service separately
docker-compose up -d --build frontend
# View service status
docker-compose ps
# Stop backend service
docker-compose stop backend
# Stop frontend service
docker-compose stop frontend- Ensure to modify environment variable configurations in
docker-compose.yml - Configure appropriate network and port mappings
- Consider using HTTPS certificates to protect services
- Regularly back up database and report files
- Install dependencies:
pip install -r requirements.txt - Run tests:
python -m pytest - Start development server:
python app/app.py
- Install dependencies:
npm install - Start development server:
npm run dev - Build production version:
npm run build
- API Keys: Ensure correct configuration of OpenRouter API and SerpAPI keys, otherwise report generation functionality will not work
- Resource Requirements: Running local Embedding models requires certain memory resources (recommended at least 8GB RAM)
- Network Connection: Academic literature retrieval requires a stable network connection
This project is for learning reference only, aiming to help researchers understand the automated process of academic report generation and related technical implementations.
- Usage Responsibility: All impacts resulting from the use of this project are unrelated to the project authors
- Compliance: Users should ensure compliance with relevant laws, regulations, and academic norms
- API Usage: When using third-party APIs, users should comply with the terms of service of the corresponding service providers
一个完整的学术论文报告生成系统,包含后端API服务和前端Web应用,支持基于人工智能的学术论文报告自动生成。
本项目旨在帮助研究人员快速生成高质量的学术论文或报告,通过集成多种学术文献检索源和大语言模型(LLM),结合检索增强生成(RAG)技术,实现从文献收集到报告撰写的全流程自动化。
- 高级学术文献搜索:集成Crossref、OpenAlex、ArXiv、Semantic Scholar等多个权威学术平台,支持并发检索和智能过滤
- 检索增强生成(RAG):融合学术文献检索结果与本地知识库,显著提升报告内容的准确性和相关性
- 智能报告生成:基于LLM的报告内容生成和组织,支持多种语言
- 本地知识库管理:支持PDF、DOCX等格式文档的加载、自动分割和向量索引
- 自定义模板功能:支持用户创建和使用自定义报告模板,灵活控制报告结构和内容
- 用户友好的Web界面:直观的报告生成和管理界面
- Docker容器化部署:简化部署和环境配置
- Python 3.9+:主要开发语言
- Flask 2.0+:Web框架,提供RESTful API
- Transformers:本地Embedding模型,支持文档向量化
- LangChain:RAG技术实现,融合检索结果与知识库
- OpenRouter API:大语言模型服务,支持智能内容生成
- 多线程并发:实现高效的学术文献并行检索
- 向量数据库:支持本地知识库的高效存储和检索
- Vue.js 3:现代前端框架
- Vite:快速构建工具
- Tailwind CSS:响应式UI设计
- Docker:容器化部署
- Docker Compose:多服务协调管理
acacademic_paper_generation/
├── backend/ # 后端API服务
│ ├── app/ # 应用核心代码
│ │ ├── api/ # API路由
│ │ ├── services/ # 业务逻辑服务
│ │ └── app.py # Flask应用主文件
│ ├── config/ # 配置文件
│ ├── database/ # 数据库文件
│ ├── requirements.txt # Python依赖
│ └── README.md # 后端服务说明
├── frontend/ # 前端Web应用
│ ├── src/ # 源代码
│ ├── public/ # 静态资源
│ ├── package.json # 前端依赖
│ └── vite.config.js # Vite配置
├── docker-compose.yml # Docker Compose配置
├── Dockerfile.backend # 后端Dockerfile
├── Dockerfile.frontend # 前端Dockerfile
└── README.md # 项目说明文档
- 克隆仓库
git clone https://github.com/S0ra-ai/academic_paper_generation.git
cd academic_paper_generation- 配置环境变量
复制并编辑后端环境变量文件:
cp backend/.env.example backend/.env
# 编辑backend/.env文件,填写API密钥等配置- 启动服务
docker-compose up -d服务将在以下地址启动:
- 前端应用:http://localhost:3000
- 后端API:http://localhost:5000
cd backend
pip install -r requirements.txt
# 配置.env文件
python app/app.pycd frontend
npm install
npm run dev- 智能文献检索:输入研究主题,系统自动从多个学术平台检索相关文献
- 知识融合:将学术文献检索结果与本地知识库内容深度融合
- 结构化报告:基于融合知识生成逻辑清晰、内容丰富的结构化报告
- 多语言支持:支持中英文等多种语言的报告生成
- 内容准确性:通过RAG技术显著提升生成内容的准确性和学术相关性
- 多源集成:同时整合Crossref、OpenAlex、ArXiv、Semantic Scholar等权威学术平台
- 并发检索:多线程并行检索,大幅提高搜索效率
- 智能处理:自动过滤、排序和去重检索结果
- 全面覆盖:确保获取最相关、最新的学术研究成果
- 标准化处理:统一不同平台的文献格式,便于后续分析和使用
- 多格式支持:兼容PDF、DOCX等多种文档格式
- 自动处理:智能文档分割、文本提取和向量索引
- 高效检索:基于语义的快速文档内容检索
- 持续优化:支持知识库的动态更新和扩展
- 安全存储:本地存储确保数据隐私和安全
- 历史记录:查看和管理所有历史生成的报告
- 内容预览:在线预览报告内容
- 便捷下载:支持报告的导出和保存
- 模板创建:用户可根据需求创建个性化报告模板
- 模板管理:支持模板的保存、编辑和删除
- 灵活配置:可自定义报告结构、章节布局和内容要求
- 模板应用:在报告生成时选择使用自定义模板
- 结构控制:通过模板精确控制生成报告的逻辑结构和内容组织
- 复用性:创建的模板可重复使用,提高报告生成的一致性和效率
后端API服务提供以下主要端点:
- GET /health - 检查API服务状态
- POST /api/generate-report - 生成学术报告
- 请求体:
{"query": "研究主题", "limit": 6, "language": "zh"} - 响应:生成的报告内容
- 请求体:
- GET /api/reports - 获取报告列表
- GET /api/reports/{id} - 获取报告详情
- DELETE /api/reports/{id} - 删除报告
- GET /api/get-templates - 获取模板列表
- POST /api/upload-template - 上传新模板
详细API文档请参考 后端服务说明。
主要配置项(位于backend/.env):
OPENROUTER_API_KEY- OpenRouter API密钥SERPAPI_KEY- SerpAPI密钥(用于某些检索功能)ACADEMIC_EMAIL- 学术邮箱(用于API请求标识)LLM_MODEL_NAME- 使用的LLM模型名称KNOWLEDGE_DIR- 本地知识库目录SAVE_DIR- 报告保存目录
前端配置主要通过环境变量和配置文件管理:
VITE_API_BASE_URL- 后端API基础URL
项目提供了完整的Docker配置,支持一键部署:
# 构建并启动所有服务
docker-compose up -d --build
# 单独构建并启动后端服务
docker-compose up -d --build backend
# 单独构建并启动前端服务
docker-compose up -d --build frontend
# 查看服务状态
docker-compose ps
# 停止后端服务
docker-compose stop backend
# 停止前端服务
docker-compose stop frontend- 确保修改
docker-compose.yml中的环境变量配置 - 配置适当的网络和端口映射
- 考虑使用HTTPS证书保护服务
- 定期备份数据库和报告文件
- 安装依赖:
pip install -r requirements.txt - 运行测试:
python -m pytest - 启动开发服务器:
python app/app.py
- 安装依赖:
npm install - 启动开发服务器:
npm run dev - 构建生产版本:
npm run build
- API密钥:确保正确配置OpenRouter API与SERPAPI密钥,否则报告生成功能将无法使用
- 资源需求:运行本地Embedding模型需要一定的内存资源(建议至少8GB RAM)
- 网络连接:学术文献检索需要稳定的网络连接
本项目仅供学习参考使用,旨在帮助研究人员了解学术报告生成的自动化流程和相关技术实现。
- 使用责任:使用本项目产生的一切影响与项目作者无关
- 合规性:使用者应确保遵守相关法律法规和学术规范
- API使用:使用第三方API时应遵守相应服务提供商的使用条款