Project Name: HiFive Goal: Help students and researchers discover personalized AI research topics, connect with relevant researchers, and explore projects through an interactive, AI-powered platform.

Mission: Guide users through a personalized journey of research discovery:

1. Assess research interests through RSTI (Research Style & Topic Inventory) test
2. Discover personalized topic recommendations based on assessment
3. Explore research areas, people, and projects
4. Connect through an AI chat assistant for deeper engagement

• RSTI (Research Style & Topic Inventory) assessment to understand user interests
• Interactive questionnaire that evaluates research preferences and topic alignment
• Generates a personalized profile based on responses

• AI-powered topic recommendations based on RSTI test results
• Curated research areas tailored to user's interests and academic background
• Visual presentation of recommended research directions

Users can explore through multiple pages:

• Overview of AI research topics and trends
• Interactive visualizations of hot topics
• Wordcloud and trending graphs

• Discover researchers and potential collaborators
• Filter by research interests, affiliation, and expertise
• View researcher profiles with publications and projects

• Browse current research projects
• See project details, team members, and research focus
• Visual cards showing project information and faculty involvement

• Interactive AI assistant for research questions
• Ask about specific topics, researchers, or projects
• Get personalized guidance based on your RSTI profile

⚙️ Tech Stack Layer Technology Frontend HTML + CSS + JavaScript (or React / Vite optional) Visualization Chart.js / D3.js / Plotly + Wordcloud.js Backend Python (FastAPI or Flask) Database None (fetch via OpenAlex API; mock JSON fallback) API Integration OpenAlex API, OpenAI API Optional Google Scholar / Arxiv scraping (Mock for demo) 📁 Project Structure ai-research-dashboard/ │ ├── frontend/ │ ├── rsti.html # Landing page - RSTI test │ ├── lootbox.html # Personalized recommendations page │ ├── research.html # Research topics and trends │ ├── people.html # Researcher discovery & profiles │ ├── projects.html # Research projects showcase │ ├── chat.html # AI chat assistant │ ├── script.js # Main frontend logic │ ├── styles.css # Global styling │ └── assets/ │ ├── images/ # Project and researcher images │ └── icons/ # UI icons │ ├── backend/ │ ├── app.py # FastAPI/Flask main server │ ├── openalex_api.py # Functions for querying OpenAlex │ ├── openai_api.py # OpenAI query + summarization │ ├── mock_data.py # fallback researcher/topic data │ ├── utils/ │ │ └── wordcloud_generator.py │ └── requirements.txt │ ├── data/ │ ├── mock_researchers.json │ ├── mock_topics.json │ └── openalex_sample.json │ ├── README.md # (this file) └── .env.example # API keys for OpenAI and OpenAlex

🔌 API Design

1. /api/wordcloud

GET → returns { "words": [{ "text": "LLM", "value": 120 }, ...] }

2. /api/trending

GET → returns { "topics": ["AI safety", "diffusion models"], "counts": [120, 98] }

3. /api/researchers

GET → parameters:

topic

institution

country Returns: JSON list of researchers with name, link, and affiliations.

4. /api/chat

POST → { "query": "AI safety" } Returns:

{ "summary": "AI safety has grown rapidly in 2025...", "suggested_researchers": [ { "name": "John Doe", "link": "https://scholar.google.com/...", "field": "AI alignment" } ] }

1. User arrives at Landing Page → Takes RSTI test by answering questions about research interests and style

2. Completion of RSTI → Receives personalized topic recommendations based on test results

3. User explores Research Page → Views trending AI topics, word clouds, and research trends

4. User visits People Page → Discovers researchers aligned with their interests, views profiles and expertise

5. User browses Projects Page → Explores current research projects and finds potential opportunities

6. User engages with Chat Page → Asks questions like "Tell me more about multimodal AI" or "Who is working on reinforcement learning?"

   • AI assistant responds with relevant information
   • Suggests specific researchers and projects
   • Provides personalized guidance based on RSTI profile

🧠 Integration Notes

OpenAlex API endpoint: https://api.openalex.org/works?filter=concepts.id:

OpenAI Model: gpt-5 or gpt-4o-mini for narrative and suggestions

Mock Mode: If no API key, load local mock JSON instead

Frontend visualization: Use D3.js for graphs, WordCloud2.js for topic cloud

Personality: Friendly, knowledgeable research guide Goal: Help users understand research trends, discover relevant researchers and projects, and provide personalized guidance based on RSTI profile

Example Interactions:

User: "What research areas match my interests?" Assistant: "Based on your RSTI profile showing interest in applied AI and human-computer interaction, you might enjoy exploring projects in multimodal learning and AI accessibility. Check out Dr. Smith's work on the People page!"

User: "Tell me about reinforcement learning" Assistant: "Reinforcement learning is trending toward large-scale agent coordination and real-world applications. I found several researchers and projects in this area - would you like me to suggest some?"

• User Profiles: Save RSTI results and track exploration history
• Advanced Matching: ML-based researcher-project matching algorithm
• Collaboration Features: Direct messaging and connection requests
• Enhanced Visualizations: Interactive co-author networks and research topic maps
• Integration Expansions: Google Scholar API, Arxiv, and university research databases
• PDF Analysis: Automatic paper summarization and key insight extraction
• Personalized Dashboard: Custom feeds based on research interests and RSTI profile

🧪 Mock Data Example

mock_researchers.json

[ { "name": "Alice Zhang", "affiliation": "MIT CSAIL", "link": "https://scholar.google.com/citations?user=abc123", "topics": ["LLM", "NLP", "AI Safety"] }, { "name": "Mark Liu", "affiliation": "Stanford AI Lab", "link": "https://arxiv.org/a/liu_m_1.html", "topics": ["Multimodal Learning", "Vision-Language Models"] } ]

💡 Prompt for Claude

Generate a full-stack web project based on this README. Use Python (FastAPI) backend + HTML/JS frontend. Integrate OpenAlex API for research data, OpenAI API for chat. Include wordcloud, trend graph, researcher filter page, and mascot chat interface. Create mock data for testing without keys.