A Review on Rule-Based and Hybrid Stemming Techniques

Reducing a word to its original form called stemming is extremely important in the field of natural language processing, especially to develop morphological analyzer and information retrieval systems. Use of stemming has decreased the dimension of indexing. Stemming algorithms are developed during last few decades to automate the process requires different approaches as it is mostly language specific. In this paper we have discussed and given a comparative view about different stemming approaches developed so far. It focuses mainly on four stemming techniques namely Rulebased, Statistical, Hybrid and Unsupervised approach.