Development and Integration of an Odia Stemmer in Dspace for Odia Search Engine

Abstract

Stemmer is used for reducing inflectional or derived word to its stem. This technique involves removing the suffix or prefix affixed in a word. It can be used for information retrieval system to refine the overall execution of the retrieval process. This process is not equivalent to morphological analysis. This process only finds the stem of a word. This technique decreases the number of terms in information retrieval system. There are various techniques exists for stemming. Here a new hybrid stemmer has developed named as “Mula” for Odia Language. It is a combination of brute force and enhanced suffix strippingapproach for Odia language. The new born stemmer is both computationally inexpensive and domain independent. We have integrated this stemmer in existing Dspace for Odia text retrieval System. The results are commendable and suggest that the new stemmer can be used effectively in Odia Search Engine. The proposed stemmer also handles over-stemming and under-stemming effectively.