Ron Miller
SemanticWeb.com Contributor

When Bing launched a few weeks ago, Microsoft made a big point of calling it a "decision engine" rather than a search engine. Helping people make decisions suggests that it might have some semantic underpinnings. You may recall that just one year ago Microsoft purchased Powerset, a semantic search company, and in fact at least some of the Powerset technology has been incorporated into Bing.

But does that mean Bing is using semantic technology? There are some semantics in play here, although it appears it's limited for now.

Decisions, Decisions

Just because a search engine claims it's better at making decisions doesn't necessarily mean it can. Mills Davis, managing director at Project10X says it could be more about marketing than technology.

"With respect to helping consumers make decisions," Davis says, "it's a novel way to market search services. Also, semantic technologies can help. However, the jury is still out as to whether Bing really delivers on this claim."

But Mark Johnson, a senior program manager in the Powerset division of Bing, says it's more than mere hype.

"In order to create a decision engine, we must have a better understanding of user queries, of Web pages, and of other kind of data to help users find the information they want," Johnson says. He adds, "With Bing, we are introducing features/functionality to help people make better decisions more quickly."

Among these are best match clearly highlighting the "best result" to a query, providing hard to find information like customer service phone numbers right on the search page when available and delivering a more organized experience.

So Where Are the Semantics?

Davis says that Powerset is clearly being used in Bing at least for natural language queries. "Powerset is a part of Bing. For example, if you ask a question, then Bing seems to drop you into Powerset, which semantically searches Wikipedia to give you an answer. But, it is not clear that Powerset is playing any wider role at this point," he says.
Johnson suggests trying a question such as "who founded General Motors" to see this functionality in action (although in testing this it's better at answering historical questions than, for example, "who owns the Boston Red Sox" or "which state has the best seafood?")

Bing uses Powerset technology to find this Wikipedia article on the founder of GM.

Johnson says that "semantics, so broadly conceived as better understanding, are woven throughout the Bing experience." He points to underlying ontologies in Bing as an example of semantics in action. "For many of the entries in the Explore Pane (left hand rail), we use an ontology," Johnson says. "If you look at the Explore Pane for a search like "Buffalo Bills," (my hometown team) you'll notice that the options in the Quick Tabs (top section of the Explore Pane) are related to the sports team."

Quick tabs like these make use of an underlying ontology related to the search query, in this case, "Buffalo Bills."

Brooke Aker, CEO at Expert System, a company that makes semantic software, sees additional semantics in Bing. "Semantics are clearly playing a role for the semantic dials you see on the left side [of the Reviewer listings] during the shopping experience," Aker says.

Semantic technology comes into play with this sentiment graph created automatically from the Reviewer information.

Bing clearly has some semantic technology under the hood, although it is scattered at this point. Whether it expands or not remains to be seen, but for now it could actually be helping its users, at least in some instances, make better decisions.