Enhancement of rules-application-order (RAO) stemming algorithm based on the first character of Malay word / Edatul Muliana Ghazalli

• Main Author: Ghazalli, Edatul Muliana
• Format: Thesis
• Language: English
• Published: 2005
• Online Access: http://ir.uitm.edu.my/id/eprint/1429/1/TB_EDATUL%20MULIANA%20GHAZALLI%20CS%2005_5%20P01.pdf; http://ir.uitm.edu.my/id/eprint/1429/
• Institution: Universiti Teknologi Mara

Stemming is important thing to improve retrieval effectiveness. Stemming is used to reduce the size of indexing file for relevancy of document retrieval. Stemming is technique to truncate the word into the root word that will reduce vocabulary size and improve recall. The Malay affixes consist of four different types such as prefix, prefix-suffix, suffix and infix. An effective and powerfiil of Malay stemmer is it just not to move the suffixes rules only but it must remove all four types of affixes. Without removing all the affixes, the stem caimot be effectively used to index of Malay documents. So in order to get the best order of morphological rule for effective and powerfiil stemmer the researcher has to find out the best order of morphological rule to stem Malay words based on first character for each alphabet. This project involves the use of two combinations simultaneously. The words that could not stem correctly by the first combination of best order which is primary will shift to alternative combination of best order of morphological rule. The resuhs of experiment B, which is enhance project is better than experiment A, which is Rules-Application-Order (RAO) by Fatimah (1995) because that algorithm has successfully stemmed all word begin with alphabet "A" until "Z" that extracted fi-om Quran documents.