The objective of this project is to develop techniques for building highly scalable and effective metasearch engines and related techniques. A metasearch engine interacts with multiple local search engines so that a single query can be used to search multiple local search engines. They study two types of metasearch engines. The first type combines multiple document search engines and will be called document metasearch engines. The second type combines multiple database driven search engines and will be called database metasearch engines. For both types of metasearch engines, the issues that they study include how to discover and classify search engines, how to build wrappers for search engines, how to identify potentially useful local search engines for each user query, and how to merge the results from multiple local search engines. For database metasearch engines, they also study how to integrate the search interfaces of multiple search engines into a unified interface and how to annotate the retrieved results. This has been added to Deep Web Research Subject Tracer™ Information Blog.
posted by Marcus Zillman |
4:25 AM
