Title :
Evaluating the weighted-keyword model to improve clinical question answering
Author :
Cao, Yong-Gang ; Ely, John ; Yu, Hong
Author_Institution :
Univ. of Wisconsin-Milwaukee, Milwaukee, WI, USA
Abstract :
Physicians ask many complex questions during their encounters with patients. Question answering systems provide immediate and direct answers to ad hoc clinical questions, and because these systems might aid in the practice of evidence-based medicine, we are developing the clinical question answering system, AskHERMES, to generate answers to such questions. In this study, we report the evaluation of a new weighted-keyword model for improving our question answering system. As part of this development, a physician manually examined AskHERMES´ answers to 20 ad hoc clinical questions created with and without the weighted-keyword model. The results show that the weighted-keyword model improves quality in question answering. AskHERMES can be accessed at http://www.AskHERMES.org.
Keywords :
information retrieval; medical information systems; AskHERMES; Information retrieval systems; ad hoc clinical questions; clinical question answering system; evidence-based medicine; weighted-keyword model; Automatic control; Cardiovascular diseases; Data mining; Frequency; Information retrieval; Libraries; Multimedia systems; Osteoporosis; Speech; Unified modeling language;
Conference_Titel :
Bioinformatics and Biomedicine Workshop, 2009. BIBMW 2009. IEEE International Conference on
Conference_Location :
Washington, DC
Print_ISBN :
978-1-4244-5121-0
DOI :
10.1109/BIBMW.2009.5332084