Penggunaan Kamus Sinonim dan Hiponim Sebagai Sumber Ekspansi Kueri Dalam Sistem Temu Kembali Informasi Berbahasa Indonesia

Abstract

In search for information, a user usually inputs a small number of words as a query. A problem of vocabulary mismatch is arisen when the system does not retrieve a document that contains relevant information but does not contain words in the query. Query expansion is used to increase the chance to retrieve potentially relevant documents by adding new words to the original query. In this work, we investigated the use of synonyms, hypernyms and holonyms as the source of query expansion in retrieving scientific papers in bahasa Indonesia. We collected papers in the field of communications science into a database and allowed it accessed through a web portal with a search facility. We used big dictionary of bahasa Indonesia (KBBI) and have appended scientific terms of communications science therein. During retrieval, the documents are sorted based on their similarity with the query words. We measured the number of retrieved documents and the relevance of documents in the top positions. Query expansion is proved to increase the number of retrieved documents by approximately 23 percent compared to retrieval without query expansion. However, the relevance of the documents on the top of the list does not change a lot.