As of the June monthly release, the Semantic Text Search algorithm can be activated on the name field for any entity type. This wider activation across entity types will expand Answers’ use of neural networks to return the result from the Knowledge Graph that precisely matches the user’s intent.
Every user might use a different query to find the same answer so it is critical to understand the semantic intent and not just match for keywords. For example, with Semantic Text Search, Answers can learn that users searching for “high top sneakers,” usually want basketball sneakers. Or in career search, a candidate may search for “recruiting” positions. Answers will detect that this user would be interested in open roles like “talent acquisition associate.”
We’d love to hear how you’re using Semantic Text Search to make your search experience smarter — so drop us a comment below!
*Note: As of the June monthly release, Semantic Text Search is only available in English and will not be activated on location-type entities.