SIMILARITY MEASURE IN ELECTRONIC TESTING AT PHARMACY DATABASE

OLUTAN JAMES OMOTAYO
Computer Science, Federal University of Agriculture, Abeokuta
March, 2015
Full text (external site)
 

Abstract

The purpose of this study is to compare the pattern matching algorithm in searching for
drugs name in pharmacy database. This work has used Jaccard similarity and cosine
similarity measure in pharmacy database and ways in which the searching can be
improved and made retrieval of relevant data easy for both trained and untrained users
especially in pharmacy. At first we have investigated string metrics and compared the
performance of Jaccard similarity and Cosine similarity coefficient by conducting an
experiment on real data from pharmacy database. According to the experimental results,
Jaccard similarity performs better to detect the order of substrings in strings than Cosine
similarity. Therefore, Jaccard similarity can be adopted in pharmacy database in
searching for drugs name.