Bing Explains Why Adaptive Search Improves User Experience

    September 18, 2011

Last week, Bing unveiled a new feature in its continued effort to improve the search experience. The feature is called “Adaptive Search” and is designed to make search results more personally relevant to users.

Have you tried Bing’s Adaptive Search? Let us know.

As Stefan Weitz, a director with Bing, explained to us, the technology understands the intent and context of each query because it looks at the user’s search history. For example, if a user typically searches for films and entertainment pieces, when he searches for “Australia,” Bing understands that he is probably not searching for the country. As a result, it would rank the 2008 movie above general information on Australia on the results page.


Weitz told us that Bing is not approaching personalization as a feature of search. Instead, the search engine believes that it shouldn’t be an option since people expect search results to be targeted and personalized to their needs.

“You should just think about personalized search as search,” he said.

While Bing is trying to make results more personal, it also wants to make sure users aren’t locked into the recently controversial “filter bubble.” This concept is essentially the concern that personalization would only return results that a searcher is familiar with and agrees with, and therefore not provide any diversity.

Although Bing takes this concern seriously, Weitz told us that it believes Adaptive Search offers a well-balanced approach for users.

“You can have personalization and serendipity, which is what really the filter bubble is saying doesn’t exist because of the personalization,” he said.

He went to say that Jamie Teevan of Microsoft Research studied this area extensively and found that personalization could actually help serendipity in some cases. According to him, the personalization of Adaptive Search is one step toward developing a human-like connection that search engines need, but have not yet been able to do.

Speaking of other search engines, there have been some that have suggested that Bing’s new approach is similar to Google’s previous query feature. Although Weitz said he has not been able to look at it closely, he did say his understanding is that it is based more on back-to-back queries. For example, if a user searches for a digital camera and the very next query he or she searches for is for Amazon, it is his understanding that Google would return Amazon’s search results for digital cameras.

Based on this inference, he said that Adaptive Search is “much broader… more complex from a computer science standpoint, but more elegant from a user standpoint.”

It’s interesting that this Adaptive Search announcement comes just after Google releases its new travel search engine, which is also similar to Bing Travel. When asked about this coincidence, Weitz made a humorous reference to the incident earlier this year when Google accused Bing of stealing their results.

“I think they’re using the same algorithm to copy our features, I guess,” he said laughingly.

Putting the humor aside, Weitz did tell us that Bing was pleased with its recent gain in market share. According to him, Bing realizes that it’s a long-term game but will continue to illustrate its commitment through new features and innovation.

“We want people to expect more from search, and if they expect more from search… we think we can actually grow the overall pie of queries that are out there, and hopefully, because we are the first to market with a lot of these innovative features, we can attract more people to those features because they are only on Bing,” he said.