Extracting meaning from the Web is a difficult undertaking. Keyword search skims the surface of contextual meaning that is locked in Web pages, Tweets and feeds. That’s where semantic search comes in. The semantic web deals with looking beyond simple links that make up the web to understand a deeper meaning and context behind that content. The Ellerdale Project, which launched in alpha this past week, is hoping to add context to search by using semantic technology to power a real-time search platform.
Ellerdale mines the real-time stream, including Tweets, RSS and the, to identify topics, messages and articles that link together based on content, not keyword. So If you looked up Sarah Palin on Ellerdale’s site, you’d see a semantic graph of related content, such as Oprah Winfrey (Palin just appeared on Oprah a few days ago), The Republican Party and John McCain.
The data on the site is mostly collected from Wikipedia, Freebase, Twitter, RSS, and by crawling the web. Ellerdale then analyzes and index the data to identify topics in text. Using this information, Ellderdale will show you the latest tweets, RSS articles and trending URLs, organized by topic. The site also analyzes trends in Tweets and feeds to display trending topics and topic clusters organized by categories (i.e. politics, sports, style).
For now you can only see topics on the site and cannot actually search for any keyword. While Ellderdale’s platform is still a work in progress, it’s already caught the eye of notable angel investor Ron Conway, who has made an investment in the startup. Startups and companies using the semantic web for search is steadily growing. Microsoft bought Powerset for $100 million to gain semantic search expertise. Hakia, NetBase, Textwise, Twine and other startups are also working on semantic search.