Crazy idea: 1. Scrape citations in user’s bibliography (can we reliably get bib codes from bibitems, or only if they use full bibtex?) 2. Use andycasey/ads to get citation trees for these articles 3. Use some ML to find most important articles that cite or are cited by the already-cited articles. 4. Suggest some useful references. :boom:
Crazy idea:
💥