Three Semantic Search Engines

Now that Google has announced that it is now incorporating semantic metadata into its search results, many are hailing a major turning point in the evolution of the semantic web. However, there are already a handful of search engines that provide their own unique takes on semantic searches of the web. While some of these search engines are designed to utilize microformats and Resource Description Framework (RDF), others may be more correctly defined as natural language search engines. Here are three of the top search engines that heavily rely on semantics when determining their search results: Hakia, SenseBot and Cognition.

Hakia

Hakia describes itself as a "general use semantic search engine" that qualifies search results based upon a combination of relevancy of the content to the query, the age of the content compared to similar sites and the overall credibility of the content's website. This allows the search to mine credible content without relying on the popularity of the site or page, a factor which Google takes into heavy consideration. Hakia is currently in its BETA stage and focusing on particular topics such as medicine, science, arts, history and similar subject matters.

SenseBot

SenseBot is another semantic search engine that is surprisingly functional with a layout that screams BETA. The landing site might look a little crude, but give it a chance before writing it off. Their motto, "a search engine that finds sense in a heap of Web pages," is reflected in the search results that delivers sentences and search terms that are very closely semantically related to the search terms. Rather than just listing links to sites, SenseBot provides the best sample sentence that triggered the search result and allows you to judge the links on those merits. Although the layout of SenseBot's search results looks considerably more cluttered than search engines like Google, I found that SenseBot reliably delivers very relevant links to pages that I had never been heard of using the traditional search engines. SenseBot also features a tag cloud in every search result that shows the terms most commonly associated with the search terms that can be very useful.

Cognition

Cognition is different from other semantic search engines in that it only searches structured textual sites such Wikipedia and the Bible. The use a a form of Natural Language Processing (NLP) that is based on the semantic mapping of the English language that Cognition Technologies has built over the course of the last 24 years. Other than Wikipedia and the Bible, Cognition primarily focuses its unique search capabilities on legal and health related texts. To learn more about Cognition, visit their site for a painfully dull video introduction or just dive in and try some searches of Wikipedia; you'll find this is a very different search tool than you are accustomed to.

Sources:

http://www.sensebot.net/
http://radar.oreilly.com/2009/05/google-adds-microformat-parsin.html
http://swoogle.umbc.edu/
http://www.cognition.com/
http://hakia.com/
http://en.wikipedia.org/wiki/Natural_language_search_engine