Microsoft talks up semantic search ambitions

Semantic search capabilities were hailed by a Microsoft official Tuesday as a better way to get users to the information they need.

During an interview and a preceding presentation, Scott Prevost, principal development manager for Bing search technology at Microsoft, touted developments in search as well as possibilities. He appeared at the Web 3.0 conference in Santa Clara, Calif.

With semantic technologies, which also are being to referred to as Web 3.0, computers have a greater understanding of relationships between different information, rather than just forwarding links based on keyword searches.

The end game for semantic search is "better, faster, cheaper, essentially," said Prevost, who came over to Microsoft in the company's 2008 acquisition of search engine vendor Powerset. Prevost is still general manager of Powerset.

Semantic capabilities get users more relevant information and help them accomplish tasks and make decisions, said Prevost.

"Traditionally in the past 10 years, search has been about query and a page. It'll become more interactive, [with] more tools for helping people refine their queries. It's really going to be more about people making real-world decisions rather than just finding some links," Prevost said.

Efforts are under way to extend semantic search to include social networking posts. "I think there's all kinds of things you want to surface," Prevost said.

"When the revolution is happening in Iran and people are using Twitter as the only form of communication, ideally, the search engine should understand what that is, understand from the social network what are the important communications and help to organize that," Prevost said. Search technology has not yet reached this stage, he said. "But we're getting there," said Prevost.

Information about anti-government protests in Iran last year was disseminated through Twitter messages.

Meanwhile, freshness of Web page remains critical, said Provost. "This has been sort of a driving aspect of search engine development for the past 10 years -- the drive to improve relevance," he said.

Bing features semantic capabilities like presenting more readable captions based on linguistic and semantic analysis of content, said Prevost.  The concept of entity extraction is leveraged in Bing, providing knowledge on phrases and what they uniquely refer to, Prevost said.