ÀÇ·á Á¤º¸ °Ë»öÀ» À§ÇÑ º¤ÅÍ°ø°£¸ðµ¨ ±â¹Ý °Ë»ö ½Ã½ºÅÛÀÇ °³¹ß ¹× Æò°¡

À¯¼ö¿µ*1), ¹ÚÈñ°æ1), ÃÖÁø¿í1)

1)¼­¿ï´ëÇб³ ÀÇ°ú´ëÇÐ ÀÇ°øÇб³½Ç

Abstract : The vector space model is widely used in the information retrieval systems. This paper represents a feasibility study of the free text information retrieval based on the vector space model in the domain of medical/health care information. To effectively use it, we evaluated various factors associated with the retrieval performance using the MEDLINE test collection known as OHSUMED. Experimental results show that the case folding, stemming, and stopwords removal contribute to the performance improvement, however the traditional weighting scheme is not effective. These observation will be useful in designing the medial domain-specific information retrieval system.
keyword : Vector Space Model, Information Retrieval, Evaluation, MEDLINE

¿¬Á¦ºÐ·ù

,


÷ºÎÇÑ ÆÄÀÏ

KOSMI2005_paper_Yoo.hwp

÷ºÎÇÑ e-posterÆÄÀÏ

Åõ°í³»¿ëÀ» »èÁ¦ÇϽ÷Á¸é "»èÁ¦"¹öÆ°À» ´©¸£½Ã°í ´Ù½Ã Á¢¼öÇÏ¿© Áֽñ⠹ٶø´Ï´Ù.

 ¾Ï È£

´ëÇÑÀÇ·áÁ¤º¸ÇÐȸ
Copyright (c) 2002 by The Korean Society of Medical Informatics