“What’s with Gary Busey’s face?”
That’s a question that Google managed to infer and interpret context out of and return a meaningful result from.
Five years ago it would have been hard to imagine a search engine turning a phrase like this into a result. I mean look at it – colloquialism at its finest!
Search engine experts (including ourselves!), have been stressing the importance of conversational search for a long time, especially since the Google Hummingbird algorithm update. It’s not until you see something so explicitly conversational that you really appreciate the intricate workings of the Google algorithm and how it is being combined with the Knowledge Vault to deliver relevant results.
We can take a deeper look at how Google’s Knowledge Vault has correctly identified numerous entities and their relationships, to highlight just how advanced a system this is evolving into. The algorithm has taken the following entities from the phrase above in order to deliver a relevant response:
- Gary Busey the person
- His face, as a part of the person
- A motorcycle accident (an event)
- The medical condition (fractured skull)
- The relationship between Gary and his face
- The relationship between Gary’s face and the event
- The relationship between the event and the medical condition
- How all this combines together to produce a human query
No wonder we keep pushing our clients to get semantic markup onto their websites – empowering Google to index and understand concepts like this much more easily is a great way to get your content more visible to users.
Conversational search is set to become even more relevant. Recent talk has been of Google merging the Google Now product into Hangouts to offer a Siri and Cortana-like companion for both voice and text conversational searches. Again, this stresses the importance of understanding how your website’s users may be searching with conversational search and helping to snap them up early in the consumer journey.