Links will open in a new window. Close it when you have finished with that site.

Semantic search engines go beyond the words to try to understand the meaning of the query. They use a variety of techniques to identify entities in the query and the content (place, names, dates), discern the context of the query, and enrich the query with related or associated terms obtained through a form of taxonomy or ontology. Natural language processing for parsing sentence syntax figures into the analysis.

DuckDuckGo - helps the searcher select the intended meaning from a list of optons. Paris - do you want the city, or a person? It can help further with suggested topics.

Hakia- analyzes the text of a document for terms that represent its meaning and associates these with terms in its repository of concept relations, thus expanding the understanding of content for that document. A query is matched to the larger understanding rather than on an exact keyword match.

Kngine - aims to bring together semantic search, semantic web, and data representation technologies. Today, it has an uncanny ability to find data (Kngine stats). On general concepts it gives a choice of meanings taken from Freebase. It can also show aspects of a topic (such as is shown for placenames).

Kosmix - utilizes some semantic analysis to organize search results by format and present related topics.

Lexxe - answers questions using natural language technology - and does so quite well. It will suggest more aspects to the query through "clusters".

Quertle - uses linguistic technology and its understanding of biomedical terms to search literature at PubMed. It looks for relationship of terms: eg caffeine and sleep. The power terms represent classes of objects: eg diseases.

Yebol - uses some semantic technologies to make sense of results. It identifies top sites, and related searches, and breaks the page into parts for web, images, twitter etc.

For more see the wiki page at HLWiki Canada.

Previous sites - Last month: Timelines in search results