Google Now Searching For Synonyms

Blink, and you might have missed it. Google’s now doing synonym searching. It’s something they quietly noted in a past Google blog post and one that comes up again today in a post that covers how Google uses experiments to improve the search interface.

Back to synonym searching in a moment. The main focus of today’s post is how small changes can have a big impact on how users interact with search results. Some before and after screenshots of various search results are shown. Most compelling to me was how by making the + symbol thicker in a "Plus Box" that allows for stock quotes to be shown for a listing, usage went up. But the post also highlights how that doesn’t mean it’s an improvement. More usage might mean people are missing other important information in favor of a new gadget. Such is the thoughtful insight that goes into measuring even little changes like this.

Midway in the post, we’re told this:

The algorithm that is responsible for the titles and snippets of result pages now highlights stems and some synonyms of the original query term. For the query [hp printer drivers] we will also return results that include and highlight the word "driver".  This sort of "stemming," as it’s called, is generally a good idea, because it helps you better identify results that match your query, but not always. Experiments of this sort help us verify (or, occasionally, overturn) our assumptions regarding changes in these algorithms.

Stemming on Google isn’t new. The company has been doing it since back in 2003. It was also highlighting stemmed words back then. If you searched for "running," and it found a page with the word "run," the word "run" would be bolded in the search listing description.

So highlighting stemmed words isn’t new. What is new is that Google now goes beyond simple stemming and does synonym matching.

For example, with stemming, a search for running might match:

• run
• runs
• runners

But it wouldn’t match:

• jog
• jogs
• jogging

This is because the word run doesn’t make up part of those words (stemming means that the word you searched for makes up the "stem" of other words).

With synonym searching — or concept searching or thesaurus searching, which are other terms used — you do a search to find matches for the original word, stem variations of the word, and words that are conceptually related to it.

Google doing searches for synonyms is a big change and one I wish wasn’t buried in the middle of this other post. It deserved highlighting on its own, or in another posts where it was mentioned, such as here in July:

It is critical that we understand what our users are looking for (beyond just the few words in their query). We have made several notable advances in this area including a best-in-class spelling suggestion system, an advanced synonyms system, and a very strong concept analysis system.

I remember reading that briefly in the middle of a trip and flagging it for follow-up. Google’s doing conceptual search now? When did that happen, I thought.

Anyway, it’s official now. I pinged Google for a bit more info, and they sent:

Yes, the highlighting of synonyms has changed. Also the degree to which we understand synonyms changes and improves, and this can affect the impact on highlighting the terms. The main point is that the algorithm does change. We mostly highlight stems, not synonyms, but the distinction is lost on most people."
 

One thing I hope will come soon after this highlighting is a flag to let people know if their search as been customized to match stems or synonyms. It can make a difference to searchers, and they may not realize it happening now.

Google Now Notifies Of "Search Customization" & Gives Searchers Control from last month covers how Google does such flagging in other cases.