AI Research Analysis Platform

An AI-powered research paper analysis platform that helps you understand credibility, detect bias, and evaluate research quality with intelligent insights.

Features

• 🔍 Search Research Papers: Search from the OpenAlex database with millions of papers
• 🤖 AI-Powered Analysis: Uses Claude API to analyze papers using a comprehensive adaptive framework
• 📊 Credibility Assessment: Evaluates papers on methodological rigor, transparency, source quality, author credentials, statistical validity, and logical consistency (0-10 scale)
• ⚠️ Bias Detection: Identifies multiple types of bias including selection, confirmation, publication, reporting, funding, citation, demographic, and measurement bias
• 📈 Key Findings Extraction: Automatically extracts research fundamentals, methodology, findings, limitations, and conclusions
• 📚 Bookmark Papers: Save your favorite analyses for later review
• 🎨 Beautiful UI: Modern, responsive design with smooth animations inspired by contemporary design patterns

Tech Stack

• Frontend: Next.js 15 with React 19
• Styling: Tailwind CSS v3
• APIs:
  • OpenAlex API (paper database)
  • Anthropic Claude API (LLM analysis)
• State Management: React hooks with localStorage for bookmarks

Prerequisites

• Node.js 18.17 or later
• npm or yarn package manager

Installation

1. Clone the repository:

cd research-analysis

2. Install dependencies:

npm install

3. Create .env.local file in the project root with your API keys:

# Copy from .env.example and fill in your actual keys
CLAUDE_API_KEY=sk-ant-...

Note: Never commit .env.local to git. Add your actual API keys only in the local .env.local file.

Running the Application

Development Mode

npm run dev

Open http://localhost:3000 in your browser.

Production Build

npm run build
npm start

Usage

1. Search Papers: Use the search bar on the homepage to find research papers by topic, author, or keywords
2. View Results: Papers appear as cards with basic information
3. Analyze: Click the "Analyze" button on any paper to run the full AI analysis
4. Review Analysis: The analysis includes:
   • Credibility score with breakdown by component
   • Identified biases and their severity
   • Key research findings and methodology
   • Research perspective and theoretical framework
5. Bookmark: Save papers you find interesting using the bookmark button
6. View Bookmarks: Access your saved papers from the Bookmarks page

Project Structure

app/
├── api/
│   ├── search/          # CORE API search endpoint
│   └── analyze/         # Groq LLM analysis endpoint
├── components/
│   ├── SearchBar.tsx    # Search interface
│   ├── ResultsCard.tsx  # Paper result cards
│   └── DetailedAnalysisView.tsx  # Full analysis modal
├── lib/
│   └── bookmarks.ts     # Bookmark management utilities
├── types/
│   └── index.ts         # TypeScript type definitions
├── bookmarks/
│   └── page.tsx         # Bookmarks page
├── page.tsx             # Home page
└── globals.css          # Global styles

API Endpoints

POST /api/search

Searches for papers using the OpenAlex API.

Query Parameters:

• q (required): Search query
• page (optional): Page number (default: 1)

Response:

{
  "papers": [{ paper objects }],
  "totalHits": number,
  "hasMore": boolean,
  "currentPage": number
}

POST /api/analyze

Analyzes a paper using the Claude API with the comprehensive adaptive framework.

Request Body:

{
  "paper": { paper object },
  "fullText": "Paper content or abstract"
}

Response:

{
  "paper": { paper object },
  "credibility": { credibility assessment },
  "bias": { bias analysis },
  "keyFindings": { extracted findings },
  "perspective": { research perspective },
  "timestamp": "ISO timestamp"
}

Analysis Framework

The AI analysis uses a comprehensive framework that includes:

Credibility Assessment (0-10)

1. Methodological Rigor (0-2.5)
2. Data Transparency (0-2)
3. Source Quality (0-1.5)
4. Author Credibility (0-1.5)
5. Statistical Validity (0-1.5)
6. Logical Consistency (0-1)

Bias Detection

• Selection Bias
• Confirmation Bias
• Publication Bias
• Reporting Bias
• Funding Bias
• Citation Bias
• Demographic Bias
• Measurement Bias

Key Findings Extraction

• Research fundamentals
• Research questions and hypotheses
• Methodology details
• Primary and secondary findings
• Limitations and severity assessment
• Conclusions and implications

Research Perspective

• Theoretical framework
• Epistemological stance
• Research assumptions
• Ideological positioning
• Author positionality
• Contextual information

Customization

Update API Keys

Create or edit .env.local with your API keys (never commit this file):

# Required for Claude API analysis
CLAUDE_API_KEY=sk-ant-...

# Optional - for future features
# OPENAI_API_KEY=sk-...

See .env.example for the template format.

Styling

• Tailwind configuration: tailwind.config.ts
• Global styles: app/globals.css
• Color scheme is defined in the Tailwind config

LLM Model

To change the Claude model, update app/api/analyze/route.ts:

model: 'claude-3-5-haiku-20241022', // Change this line

See Anthropic documentation for available models.

Performance Considerations

• Papers are fetched on-demand from OpenAlex API
• Analysis requests are sent to Claude API (may take a few seconds)
• Bookmarks are stored in browser localStorage
• Consider implementing pagination for large search results

Limitations

• Analysis quality depends on Claude API capabilities
• Full-text document analysis is available when full text is provided
• Bookmarks are stored locally in the browser (not synced across devices)
• OpenAlex API may have rate limits

Future Enhancements

• Full-text PDF parsing for complete paper analysis
• Database backend for persistent bookmark storage
• User accounts and authentication
• Advanced filtering by bias level, credibility score, etc.
• Comparison view for multiple papers
• Export analysis as PDF
• Community annotations and ratings
• Advanced search filters
• Research collaboration features

Troubleshooting

API Keys Not Working

• Verify the API keys are correct in .env.local
• Check if the APIs are accessible from your network
• Try regenerating the API keys from the provider dashboards

Slow Analysis

• Claude API analysis can take 10-30 seconds per paper
• This is normal behavior for comprehensive analysis
• Consider implementing a progress bar for better UX

No Search Results

• Try simpler search terms
• Check the OpenAlex API is responsive
• Verify you have internet connectivity

License

MIT License - feel free to use this project for personal and commercial purposes.

Contributing

Contributions are welcome! Feel free to submit issues and pull requests.

Support

For issues and questions, please open an issue on the GitHub repository.