NLP resources at NLM

"September 2016 CHQA-Corpus-1.0 A collection of 2,614 consumer health questions annotated with named entities, question topic, question triggers, and question frames." That is how NLP resources at NLM of Bethesda, MD introduces itself. Measured against the twelve signals AI engines rely on, it scores 7.8 out of 10, a strong showing.

On the page it is passable on partial sitemap and link coverage. What holds it back is a heading structure that lays out what it offers, the technical files AI engines look for, such as llms.txt or an open robots.txt, structured data describing the business, a clearly stated business identity and credibility markers like credentials or reviews.

Off the page, other sites link to it, the domain reads as established, AI engines recognise it as a distinct business, it turns up in press coverage, people discuss it in places AI engines read and it comes up on Reddit.

AI crawlers have visited 3 times in our tracking, including GPTBot (ChatGPT) and ClaudeBot (Anthropic). It does link out to several social profiles, which helps engines tie the brand together.