A System of User-Guided Biological Literature Search Engine Meng Hu and Jiong Yang   View Paper (PDF)   Return to December 2005, Searching and Mining Literature Digital Libraries, Volume 28, Number 4 Abstract Efficiently finding most relevant publications in large corpora is an important research topic in information retrieval. The number of biological literatures grows exponentially in various publication databases. The objective of the study in this paper is to fast locate useful publications from large biomedical document collections based on users' preferences. In this paper, a new iterative search paradigm is introduced which integrates biological background knowledge in organizing the results returned by search engines, and utilizes user feedbacks to filter irrelevant documents. A term weighting scheme based on Gene Ontology is introduced to improve similarity measurement of documents in biomedical domain. A prototype text retrieval system has been built based on this iterative search approach. Experimental results show that the system can filter a large number of irrelevant documents while keep most of the relevant documents with limited user interactions. ©2006 Association for Computing Machinery