We’ve become so comfortable with and accustomed to (some might say unnerved by) Google’s many search algorithm tweaks these days that it has become customary to keep our eyes peeled for the next big thing.
For some search watchers, that would be semantic search, which uses XML and RDF data from semantic networks to disambiguate search queries and Web text to improve search results. Semantic search has been marginally popularized by Powerset, which Microsoft pounced on last summer, Hakia, Yahoo and others.
It seems Google has joined the semantic search party, though when exactly this happened is officially a mystery, unearthed in part by ReadWriteWeb’s Marshall Kirkpatrick, who blogged about the possibility that Google has been sneaking disambiguated search under our noses.
Marshall executed such queries as “what is the capital city of Oregon?” and “What is Britney Spears’ mother’s name?” noting:
“The answers to these and other factual questions are now displayed above natural search results in Google and the information is structured in the traditional subject-predicate-object format, or “triples,” of semantic Web parlance. The answers aren’t found structured that way on the Web pages they come from – Google appears to be parsing the semantic structure from semi or unstructured data.“
Marshall acknowledges that this in no way ensures Google is parsing with semantics and took grief over the issue by readers who claimed to have seen such results to questions for a few years. But his sleuthing work raises another interesting question about Google: If Google was doing semantic search, how would we know for sure?
I wanted to know for sure, so I asked Google for clarification today and a spokesperson told me:
“The feature noted in the ReadWriteWeb post this morning is something we first shipped in 2005. See original announcement here. It’s worth noting, this feature may have become more noticeable in the last few days because we updated it with more information from the Web. As always, Google continues to strive to understand documents and queries better to provide our users with the most relevant search results.“
“Updated it with more information from the Web” reads like Google speak for “we’re now doing semantic search.” Google won’t cop to it, though.
So Google has been offering Q&A results since 2005, but Marshall is spot on that answers to some of his queries have changed, thanks to Google leveraging disambiguation technologies not unlike what we’ve come to see in Wikipedia.
Google’s Kevin Marks, who evangelizes for the company’s OpenSocial group, weighs in:
“You may be seeing multiple different results-enhancing features – remember, Google runs 50 to 200 experiments at once, and lots of those make it through, if they enhance certain kinds of queries. That’s also why it can be hard to answer questions about this ‘officially’ as Chris Saad asks, as there are a lot of separate heuristics running that overall showed up as better results.“
So, net-net Google Q&A has been around for a few years, but we’re just beginning to see the fruits of Google’s semantic work blending seamlessly with the company’s search engine.
Just search as usual at Google.