Semantic Search Engine

Back to the days in the university, semantic web was a computer science research area and now we can see some nice search applications out there: Powerset / hakia / deepdyve

Compare the results of searching"what is software design" in Powerset , hakia and Google

What is semantic search? Wikipedia says, "Semantic search is a process used to improve online searching by using data from semantic networks to disambiguate queries and web text in order to generate more relevant results. (more)"

Here's a very good summary from Atos Consulting:

Concept

* The Semantic Search Engine or Third-Generation Search Engine is a search engine in which Semantic technologies and / or Natural Language Processing (NLP) techniques are applied. There are some investigation groups and companies―such as Swoogle, PowerSet (acquired by Microsoft in 2008), and hakia ― that are making a big effort in this area.
* Basically, the objective of these products is to take advantage of the existence of structural data (RDF, microformats, and OWL) on the Web in order to process the information included and offer to the end user an intelligent response. Essentially, the Semantic Search Engine can understand the question and is able to provide an exact response, rather than just returning thousands of pages where all words included into the search field are found.
* This challenge is so complex that most current providers cannot properly be considered as true Semantic Search because results are not oriented to the casual end user, but as a tool for specialists.
* Another related concept is Natural Language Search in which one can ask a question using complete sentences phrased in everyday language (such as ‘where can I find information about William Shakespeare?’) as opposed to formulating a search statement (such as ‘su:Shakespeare, William’).
* There is a big dependence between this technology and Semantic annotation. Until all Webmasters, blog users, and wiki writers frequently use ontology annotation, Semantic Search will be limited to certain Webpages or content provided by Semantic experts.
* Another important dependence to consider is with the language itself. Some languages (such as English) are more suitable than others in making the adoption process less complex.

Application and Benefits

* Different applications of Semantic Search are: Vertical Semantic Search and General-purpose Semantic Search. General-purpose actual search engines work using statistical ranking methods, and the results are reasonably acceptable, although the end user’s time may be wasted by search results that don’t distinguish between credible contents and misleading information. As hakia points out, a popular Website may not always be credible and a credible Website may not always be popular.
* Another application would be the ability to search precisely in non-textual content, such as annotated photograph and audio.
* Related with Vertical Semantic Search, this technology is being adopted where the meaning of sentences, the combination of content, and context are crucial to retrieve precise information, such as in law, finance, science, and libraries.