Intelligent Search Engine: Simulation to Implementation

World Wide Web (hence forth referred as web) has enveloped all spheres of life. Exponential growth in the web-users and data vouch for its popularity and demand. Search for required information from the tera bytes of data invariably reminds us needle search in a haystack. Yet it has become inevitable and essential. Hence, Search Engines (SE) are forced to cope with the web dynamics by improvising their infrastructure (both hardware and software) to yield current and concise results. Recency and Relevancy still elude the current generation search engines. In this paper we present architecture and design specifications for new generation search engines highlighting the need for intelligence in search engines and give a knowledge framework to capture intuition. We also characterize relevancy on a per user basis by quantifying user’s knowledge quotient (Kq) and user’s domain expertise (Kf). Simulation methodology to study the search engine behavior and performance is described. Simulation studies of our proposal are conducted using fuzzy satisfaction function and heuristic search criterion after modeling client behavior and web dynamics. We also share some of our implementation experiences in building SARVAGNA, a new generation search engine.