Paper Impact Analyzer

Analyze academic paper impact using multiple data sources (arXiv, GitHub, OpenAlex, Semantic Scholar). Input an arXiv ID and get a multi-dimensional impact assessment with graceful degradation when APIs are unavailable. Use when evaluating paper influence, comparing papers, or assessing whether a paper is worth reading.

Multi-source, fault-tolerant academic paper impact analysis.

When to Use

• Evaluating a paper's academic influence or community adoption
• Comparing impact across multiple papers
• Deciding whether a paper is worth reading based on external signals
• Checking GitHub stars, citation counts, venue acceptance for a paper
• Assessing author credibility (h-index) for a paper
• Batch-analyzing papers in a survey or literature review

How to Use

Single paper

Run the analysis script with an arXiv ID:

python scripts/analyze.py 2603.04948

Multiple papers

Pass multiple arXiv IDs separated by spaces:

python scripts/analyze.py 2603.04948 2602.15922 2603.05488 2602.22661

Output

The script prints a structured Markdown impact report for each paper, including:

Dimension	Example
Publication date	2026-03-05 (20 days ago)
Venue acceptance	ICLR 2026
GitHub repo	2,263 stars / 214 forks
Citation count	12 (OpenAlex) / 15 (S2)
Author h-index	First author h=23
Affiliations	UC Berkeley, UT Austin

Plus a synthesized overall rating (S/A/B/C/D) with confidence level and data completeness.

Data Sources (Priority Order)

1. arXiv API — paper metadata, authors, abstract (always available)
2. GitHub API — repo stars, forks, issues (most reliable external signal)
3. OpenAlex API — citation count (free, no API key needed)
4. Semantic Scholar API — citations, influential citations, author h-index (rate-limited)

Each source fails independently. The script always produces output using whatever data is available.

Design Philosophy

• Graceful degradation: Every API call is wrapped in try/except with timeouts. If Semantic Scholar returns 429, the report still includes arXiv + GitHub + OpenAlex data.
• Age-aware scoring: Papers < 3 months old are scored primarily on GitHub + venue + team. Papers > 1 year old are scored primarily on citations.
• No API keys required: All data sources used are free and keyless.
• Single file: The entire implementation is in scripts/analyze.py with zero external dependencies (stdlib only).