Semantic search engines...

As we keep hammering in the last months, semantics are said to be ‚ the next big thing‚ in search engine technology. Blogger huff and puff about the new experience, the changes it will bring to the eager masses. But what makes it so different from classic search engines?

First we need to check what is semantics, as stated by Wikipedia:

Semantics is the study of meaning, usually in language.

So semantic search seeks to make sense of search results based on context, by taking into account contextual meaning and  searcher intent in the search space, whether it is a personal document search system or a web search, to provide premium relevant results. It identifies concepts structures in the texts. In a search by “car”, documents containing the words like “fuel”, “wheel”, will be returned, even if car is not mentioned in document.

Google, Baidu and Bing continue to control the search engine realm, they try to match the search queries to balanced results (relevancy to context of search, while depending on simplistic traditional ranking systems). Using a semantic search engine will ensure more relevant results based on the ability to understand the definition of the word or term that is being searched for, rather than on numbers. Semantic search engines are able to understand the context in which the words are being used, resulting in smart, relevant results, more in line with the human way of thinking.

But how does it refines itself?

• It can offer a combination of automatically placed pages into dynamic categories, tags, related topics and keywords to narrow the search successfully. With a keyword like “car”the search engine would offer you a list of vehicles, with related links to brands, auto related webpages, even automobile fixing and modification. Knowing what topic a page relates to is invaluable for returning relevant results.
• Use of statistical analysis, to match keywords and their semantic, allowing the semantic information to be associated with new topics.

Semantic search is still in its infancy, Adsense already provides semantic advertising, using latent semantic indexing to enable adverts targeted to the theme of a webpage. LSI refers to how a search engine actually indexes terms and relates it to other terms that mean the same thing. In other words it will return pages that contain information that is semantically the same as the keyword searched for.

Still it already provides a basic step at the bottom of the search engine, alternatives based upon the initial search.

The search engine don’t have problems with recall and precision, but semantic search engines allow the content of a page to decide its utility. They aim to get less spam, of more relevant ads and more focused information. Engines like Yebol, Hakia and Swoogle already provide a basic service.

About these ads