Research Assistant

Overview

Research Assistant is a command-line tool that automates academic research by combining arXiv paper discovery with Large Language Model (LLM) analysis. Through an interactive CLI interface, researchers can explore topics, analyze papers, and discover new research directions with real-time progress tracking.

Features

• Interactive CLI: User-friendly command-line interface with menus and progress tracking
• Real-time Progress: Visual feedback on search, analysis, and saving phases
• Multi-LLM Support: Compatible with multiple LLM providers (Claude, GPT-4, LLaMA, Gemini)
• Concurrent Processing: Efficient parallel processing of papers and research topics
• Research Synthesis: Generates topic summaries and identifies new research directions
• Result Export: Structured saving of research results and analyses

Installation

Prerequisites

• Python 3.8+

Required API Keys

Create a .env file in the project root with at least one API key:

CLAUDE_API_KEY=your_claude_key
OPENAI_API_KEY=your_openai_key
LLAMA_API_KEY=your_llama_key
GEMINI_API_KEY=your_gemini_key

Setup

1. Clone the repository

git clone https://github.com/yourusername/research-assistant.git
cd research-assistant

2. Install dependencies

pip install -r requirements.txt

CLI Usage

Starting the Interface

python interface.py

Main Menu Options

1. Start New Research: Begin a new research topic
2. Settings: Configure tool parameters
3. Exit: Close the application

Research Flow

1. Enter your research topic when prompted
2. Watch real-time progress through three phases:
   • Search Phase: Finding and filtering relevant papers
   • Analysis Phase: Processing paper content
   • Saving Phase: Storing results

Viewing Results

After analysis completes, you can:

• Save individual topic analyses
• View detailed results
• Continue research on generated topics
• Start new research
• Modify settings

Settings Configuration

Through the CLI, you can adjust:

• Number of research topics (default: 5)
• Maximum papers per topic (default: 3)
• LLM provider (based on available API keys)
• Save directory for results

Progress Display

The CLI shows real-time progress with:

• Color-coded progress bars for each phase
• Current operation status
• Paper and topic tracking
• Detailed status messages

Requirements

arxiv==2.0.0
anthropic==0.8.1
openai==1.12.0
google-generativeai==0.3.2
PyMuPDF==1.23.8
rich==13.7.0
python-dotenv==1.0.1
llamaapi==0.1.38

Output Structure

Research results are saved in a structured format:

research_results/
├── topic_name_timestamp/
│   ├── metadata.json
│   ├── topic1.json
│   ├── topic2.json
│   └── ...

Contributing

Contributions are welcome! Please follow these steps:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Credits

Inspired by Automated-AI-Web-Researcher-Ollama

Troubleshooting

Common Issues

1. API Rate Limits: Adjust tokens_per_minute in settings if hitting rate limits
2. Memory Usage: Reduce max_papers_per_topic for lower memory consumption
3. PDF Processing: Ensure proper permissions for paper downloads

Debug Logs

Logs are stored in the logs/ directory:

• llm_search.log: Paper discovery logs
• llm_analyzer.log: Analysis process logs
• research_assistant.log: General operation logs