An innovative lemmatization technique for Bangla nouns by using longest suffix stripping methodology in decreasing order

In this proposed work, an attempt is made to find out the root part from inflected Bangla nouns by applying an innovative technique by using longest suffix stripping methodology in decreasing order. The test data is generated from a Bangla text corpus developed in the TDIL Project of the Govt. of India. The exhaustive suffix list obtained from the research work carried out at Linguistic Research Unit of Indian Statistical Institute, Kolkata while the Bangla non-inflected noun list used in this work is obtained from a wordlist generated by Pashchimbanga Bangla Akademi, Kolkata and available in the net. The algorithm is applied on randomly selected 1273 noun instances and accuracy is achieved around 94%.