Skip to content

Siri, what is semantic search?

Big Bang Theory's Raj meets his dream voice-activated "personal assistant", Siri


I keep reading bits and pieces about Google using semantic search technology. It's a way that they plan to answer user questions rather than simply hunt down words. For quite awhile I had read about this being the next big thing in search engine functionality. I'm no expert on search technology but it seems that most of the searching we now do is semantic.

We have become used to seeing our search results as list of blue links. That's how it has been since the beginning. But you are probably noticing more things on the top of that results page. The promise of semantic search is to improve search accuracy by understanding our intent and also the contextual meaning of the terms we search.

Although Apple's Siri has both fans and critics, the potential is pretty exciting. It can support natural language now and it will get better at semantic search in the future. Instead of just using a keyword-based search, we can ask a question. Instead of "restaurant, Italian and a city", you ask "Where's a good place to get Italian food nearby?" Is that question really so much easier than the keyword search? Well, Siri. Alexa or Google or the next big thing in search will know where you are located already (via GPS or your IP address) and it will also know that we have been to other restaurants in the area. And if we have entered some social data, like reviews of those restaurants, it will be able to suggest somewhere to eat that fits out personal little algorithm.

At least that's is the plan.

It was a year ago that I read an article about Google's semantic search algorithm. and since making it work requires vast amounts of data, Google seems like the one to do it.

I have also read that "true" semantic search uses an "inference engine" which means that - like a good human reader - instead of just recognizing words, it draws on its own "knowledge" to reach a conclusion.  Hello, ontology!

Something else that needs to happen is that users need to change our searching habits. We will have to move from basic keywords that result in thousands of results, to clearer queries that recognize the capabilities of the engine. We are not at the point where you can ask "What would I like for dinner tonight?" and expect an answer.

Trackbacks

No Trackbacks

Comments

Display comments as Linear | Threaded

No comments

The author does not allow comments to this entry

Add Comment

Enclosing asterisks marks text as bold (*word*), underscore are made via _word_.
Standard emoticons like :-) and ;-) are converted to images.
BBCode format allowed
E-Mail addresses will not be displayed and will only be used for E-Mail notifications.
To leave a comment you must approve it via e-mail, which will be sent to your address after submission.
Form options