Is Microsoft stalking Powerset's search technology?
While Powerset is preparing for the public rollout of its unique, semantic search engine, Microsoft may be interested in acquiring the start-up, according to sources.
I asked Barney Pell, Powerset co-founder and CTO, whether there was any truth to a Microsoft-Powerset deal rumors. He said, "No comment," and noted his policy of not commenting on rumors. Microsoft also declined to comment on rumors.
Powerset co-founder and CTO Barney Pell
(Credit: Dan Farber)Bringing Powerset, which has no revenue and a tiny user base at this point, into the fold would be spare change for Microsoft compared with spending $45 billion to $50 billion on Yahoo. But, it could bring something useful to Microsoft--and Yahoo, if their union were consummated--in the battle for search users with arch rival Google.
Powerset raises the bar on search based on a preview that I had of the service last month. Powerset differs from the Google in that it extracts and indexes concepts, relationships, and meaning, rather than keywords. It's able to create connections and pivot in some cases in ways that elude Google's proficient engine, which favors more of a statistical approach.
Powerset uses a sophisticated natural language parser (licensed from Xerox PARC) to find subjects, verbs, objects, synonyms, and other elements for indexing.
Initially, Powerset is performing its magic on the 3 million pages of Wikipedia content, enabling a new kind of search and navigation experience on the popular information resource.
A next step would be to index the Web, which would be of great interest to Google rivals. Powerset has garnered $12.5 million in Series A funding from Foundation Capital, Founders Fund, and angel investors. Given the cost to scale up a semantically rich index of 20 billion Web pages, Microsoft would be a good match for Powerset. Then again, so would Google. Stay tuned...
Dan Farber is editor in chief of CBS Interactive News, which includes CBSNews.com and CNET News. He has more than 25 years of experience as an editor and journalist covering technology. E-mail Dan. 
- by Yegor Kuznetsov May 13, 2008 7:14 AM PDT
- Dan, great story on Powerset, thank you very much!
- Reply to this comment
-
(3 Comments)One thing confuses me, though.
According to you posting, ?Powerset differs from the Google in that it extracts and indexes concepts, relationships, and meaning, rather than keywords.? Then you write that ?Powerset uses a sophisticated natural language parser (licensed from Xerox PARC) to find subjects, verbs, objects, synonyms, and other elements for indexing.?
Subjects, objects, verbs and synonyms are still keywords, are they not?
It looks like Powerset?s engine still relies on exact definitions, although it processes them in a unique way.
A really ?keyword-free solution? does not rely on exact definitions at all. For example, Brainware?s Globalbrain (recently found by Redmond magazine to be superior to Google and Microsoft desktop search offerings) differs from other search products on the market in that it does not make use of keyword search.
Globalbrain?s searches are natural language based, using a patented n-gram approach. While indexing the data in preparation for search, it breaks words and phrases into successive three-letter groups. It then breaks up the search term in the same manner, and performs comparisons of its three-letter groups to those in the index. Then Globalbrain rates the possible hits based on matches to those three-letter groups.
The capability provided by Brainware lets users search, not simply on key words, but on whole paragraphs and pages. Also, it is language-independent.
Here?s a case study showing Brainware in action:
Fulbright & Jaworski: Leading Law Firm Searches And Shares Knowledge Base Smarter, More Accurately
http://www.brainware.com/brain_case_lawfirm.php