A Light Weight Stemmer for Urdu Language: A Scarce Resourced Language

Stemming is a procedure that conflates morphologically related terms into a single term without doing complete morphological analysis. Urdu language raises several challenges to Natural Language Processing (NLP) largely due to its rich morphology. The core tool of information retrieval (IR) is a Stemmer which reduces a word to its stem form. Due to the diverse nature of Urdu, developing its stemmer for an IR system is a challenging task. This paper presents a light weight stemmer for Urdu text, which uses rule based approach. Exceptional lists are developed to enhance the accuracy of the stemmer. The result of the stemmer is quite enough and can be effective in IR system.