ResearchScope

A static-first research intelligence platform for CS papers — track what matters, who drives it, what to read first, and where the research gaps are.

---

What is ResearchScope?

ResearchScope is an open, static research intelligence dashboard for computer science papers. It is not primarily a CLI tool — it is a website that is rebuilt daily by a GitHub Actions pipeline and published to GitHub Pages.

The pipeline fetches papers from arXiv and ACL Anthology, enriches them with scores and tags, detects research gaps, and writes the results as static JSON + HTML to the site/ folder. The site then renders everything from those JSON files in the browser — no server required.

---

Features

Feature	Description
📄 Paper intelligence	Papers from arXiv & ACL Anthology scored by recency, citation momentum, and relevance
👩‍🔬 Author / lab intelligence	Track prolific authors and their recent output
🗺 Learning paths	Curated reading paths tagged by topic and difficulty level
🔍 Research gap explorer	Surface under-explored areas and emerging directions
🖥 Static dashboard	Zero-backend site hosted on GitHub Pages, updated daily

---

Data Sources (MVP)

Source	Content
arXiv	Preprints across CS, ML, NLP, CV, and related fields
ACL Anthology	Peer-reviewed NLP / CL papers from ACL, EMNLP, NAACL, and others

---

Architecture

┌─────────────────────────────────────┐
│          GitHub Actions              │
│  (runs daily via cron schedule)      │
│                                      │
│  src/pipeline.py                     │
│    ├── connectors/   arXiv + ACL     │
│    ├── scoring/      read-next rank  │
│    ├── tagging/      topic tags      │
│    ├── difficulty/   reading level   │
│    ├── gaps/         gap detection   │
│    └── sitegen/      → data/*.json   │
└──────────────┬──────────────────────┘
               │ commits data/
               ▼
        site/  (GitHub Pages)
          index.html   – dashboard homepage
          papers.html  – full paper list
          authors.html – author profiles
          gaps.html    – research gap explorer
          topics.html  – topic browser
          labs.html    – lab profiles
          assets/      – CSS + JS
          data/ (symlink / copied) ← JSON from pipeline

Project Layout

.github/workflows/update.yml   # daily pipeline + Pages deployment
src/
  pipeline.py                  # orchestrates all stages
  connectors/                  # arXiv & ACL Anthology fetchers
  scoring/                     # read-next scoring
  tagging/                     # topic tagging
  difficulty/                  # difficulty assessment
  gaps/                        # research gap extraction
  sitegen/                     # writes data/*.json for the site
site/
  index.html                   # homepage (published to GitHub Pages)
  papers.html / authors.html … # sub-pages
  assets/css/ assets/js/       # static assets
data/                          # generated JSON (committed by CI)
tests/                         # pytest test suite

Local Development

# Clone and install
git clone https://github.com/kishormorol/ResearchScope.git
cd ResearchScope

python -m venv .venv
source .venv/bin/activate    # Windows: .venv\Scripts\activate
pip install -r requirements.txt

# Run the pipeline locally (writes JSON to data/)
python src/pipeline.py

# Run tests
python -m pytest tests/ -v

# Open the site locally (serve site/ with any static server)
cd site && python -m http.server 8080

GitHub Pages Deployment

The workflow in .github/workflows/update.yml:

1. Runs the Python pipeline to fetch and process papers.
2. Commits any changed JSON files in data/.
3. Uploads the site/ folder as a GitHub Pages artifact and deploys it.

To enable Pages for a fork: go to Settings → Pages and set the source to GitHub Actions.

Contributing

See CONTRIBUTING.md for guidelines.

Acknowledgments

ResearchScope aggregates metadata from the following open academic resources. We are grateful to each for making their data publicly accessible.

Source	What we use	License
arXiv	Paper metadata (title, abstract, authors, categories)	Metadata: CC0 (public domain)
ACL Anthology	NLP/CL paper metadata via anthology+abstracts.bib.gz	2016+: CC BY 4.0 · Pre-2016: CC BY-NC-SA 3.0
PMLR	ICML proceedings metadata	CC BY 4.0
Semantic Scholar	Conference paper metadata and author affiliations	See S2 API License

ResearchScope stores only bibliographic metadata (titles, abstracts, authors, affiliations, URLs). No full paper text or PDFs are stored or redistributed. All papers link back to their original sources.

If you are a publisher and have concerns about how your content is represented, please open an issue.

License

MIT © 2026 Md Kishor Morol