Development of SearchBadger, A Framework for Evaluation of Search Results

摘要

Google represents a significant share in the search market and dominates the world of search engines. Although Google has constantly made effort to improve its search results by adding a result page preview and offering a block URL option, individual user feedback to the real time search operation is still very limited. In this paper, a new framework called SearchBadger is proposed to optimise search results in real time by interacting with user and taking into account immediate user feedback. The proposed framework SearchBadger is developed using C#, db4o and Visual Studio. SearchBadger architecture contains the major component BadgerWorkspace and is followed by graphical user interface and search behaviour. Keyterm significance and stemming are discussed and then page re-ranking is achieved by instant user feedback. The framework can be integrated to the existing search engines or served as a test bed for future development of semantic web.