Query Expansion on Information Retrieval System in Indonesian Language using Thesaurus

Abstract

Query expansion is a common technique for fixed query in order to yield a better information retrieval result. Using a thesaurus is a way for fixed query. Thesaurus is a dictionary of words and terms which are collected and filtered by correlation of meaning of the word, like synonim, antonim, hyponim, and hypernim. This research is intended to analyze the effectivity of query expansion using thesaurus Global Analysis Method. There are a thousand agricultural documents and 93 medical plants documents. The research used Tf-Idf as a weighting method. The result of system performance which were used in this research is good with average precision of 53% for agricultural document and 76% for medical plants document using thesaurus query expansion. The maximal result that the thesaurus could gain is by using two-word length query. Query expansion method using thesaurus has a better result than the one using 2-language dictionary and pseudo relevance feedback document segmentation. The query length and the taking of some number of word group in thesaurus will affect the result of information retrieval.