From Distributed Information and Intelligence Analysis Group

Jump to: navigation, search

The rapid growth of the internet has gone beyond placing information at our fingertips. In fact, the explosion of information (not just text, but also images, and videos) now flies under our fingertips so fast we cannot even feel them. The information space is growing and changing so rapidly that it is a challenge trying to keep the user up-to-date with the information he really needs. Existing information systems uses keyword mapping mechanism, which focuses on rapidly sorting information but is still ignorant of information content and, most importantly, relevance to the user. Our solution to high quality information retrieval is called I-FGM. It is a unified system to retrieve heterogeneous information from massive and dynamic search spaces. Using I-FGM, users can retrieve textual and image information together by one simple click. One important feature of I-FGM is that it analyzes the semantic and syntactic meaning of the input query and the documents in the search spaces so that it understands the contents before judging the relevance of the information. Moreover, since the search space is dynamic, IFGM automatically updates the information and always displays the most recent result in a rolling fashion. A prototype of I-FGM has been built on a cluster of 100 high-end computing nodes. It has a user-interface that is available to the user via Internet and can handle multiple queries from multiple users simultaneously. PARTNERS: Virginia Tech, Wisconsin-Whitewater

Personal tools