Hybrid Search Engine: Concept and Framework

The web allows people to share the information available on the websites through generically known search engines. Google, the most popular search engine, is not the end all and be all of search engines. There are actually some better and more efficient search engines out there, and there are different types of search engines worthy of information seeker's attention. The mostly influencing factor in the searched content is a portion of the web; Surface or Deep. More than 80% of the entire web is part of the deep web and this area does not allow to be searched by general search engines like Google and Yahoo. Therefore, a large amount of valuable resources are definitely missed out. The entrances of the deep web are so called vertical search engines which focus on a specific area or a segment of interest. The best thing to explore the hidden web is to consult with a higher number of search engines as possible, since no search engines reach all the information available on the entire web. In this paper, we suggest a new type of a search engine that works with general and vertical search engines together in parallel processing. In doing so, the proposed search engine grafts specialty and accuracy from vertical engines and a wide range of data from general engines. This paper highlights the main requirements for a preliminary design of the suggested search engine, compares our suggestion With federated search engines, and considers research agendas that need to be addressed before implementing.