Powerset, the San Francisco-based search engine startup, plans to focus on plain language-based search as opposed to keyword-based search prevalent among major search engines today. The company offers users a peek at what to expect from the natural-language engine.
During the initial run, Powerset will index pages from Wikipedia (NYTimes), and it will require users to vote on the relevance of articles returned in comparison to Google. There is significant hype around the search engine, which has helped it gain $12.5 million in venture backing (BizJournal).
A quote from the article at Associated Press:
This isn’t the first time a search engine has tried to understand simple English, but Powerset has drawn more attention because its natural-language technology is being licensed from the Palo Alto Research Center.
Better known as PARC, the Xerox Corp. subsidiary is renowned for hatching breakthroughs — like the computer mouse and the graphical interface for personal computers — that were later commercialized by other companies.
Natural-language search has been tried before, but it hasn’t been as successful as keyword-based searches. Thus, Powerset has its batch of skeptics (Business Week) who feel that apart from the controlled environment in which Powerset is doing well, the real test is in real-world query analysis and relating the innumerable combinations of queries to relevant documents.
There are several other natural-language engines, such as Hakia. However, the power behind Powerset seems to be the initiative to let users have a voice in the crafting of the technology.
To pose a significant challenge to Google, Powerset has to come up with a compelling offer that Google cannot up with its own extensive research in the field of semantics and natural language.