Paper to HTML
lastmod 2021-09-18created 2021-09-181 edits
Converts papers into easy-to-read HTML
Features
- Uses machine learning to read the content of PDF, LaTeX, and PubMed Central XML and convert it to HTML
- Aimed at improving accessibility
- An experimental prototype of Semantic Scholar, an AI-based research search engine
- Currently caches only the extracted images/content, used solely to serve people who upload the same document faster. Uploaded files are not kept
- Limitations
- Tables are extracted as images
- Math content has low accuracy or is barely extracted
- LaTex/PubMed processing may lack some features compared to PDF