Semantic search
Search that matches the meaning of a query, not only the exact words in it.
What it means
Keyword search looks for the words you typed. Semantic search turns text into numerical vectors, so a query about heart failure readmission can find a paper that discusses rehospitalization after heart failure. Many academic search tools combine both methods, because exact terms still matter for names, codes and technical phrases.
Semantic search is useful for research because the best paper on a topic often uses different words than your query. It can also miss results that depend on one precise term. Try both kinds of search on the same question and compare the top results. Semantic Scholar's public API, for example, exposes paper search and embeddings that developers can build on. For a single search, start with a few query phrasings and compare how the top results change.
Continue this reading path
Cite this entry
https://bestaiforresearch.com/glossary/semantic-search/