This is a Gujarati stemmer in Java. Stemming is a process in which affixes are removed form the root word (stem). It relates morphological variant words to corresponding common root. For example "પ્રતિઉપયોગી" is word which has stem " ઉપયોગ". Stemmers are language specific tools. The design of a stemming algorithm requires a significant level of linguistic expertise. There has been lot of significant work in the development and evaluation of stemmer for non-Indian languages, but very less or no significant work has been done on Indian front especially for Gujarati language.The code of this stemmer is based on algorithm designed under guidance of Prof. Nikita Desai, India. It takes input file of type .txt containing Gujarati text encoded as UTF-8 and then removes stop words which are unessential. After processing rest of the words, it outputs corresponding file containing all stem words plus other details.
Features
- Offline stemmer for Gujarati language which is one of Indo-Aryan language used in India.
- Separate class to check for prefix, suffix, substitution & for dictionary check.
- Can take any file containing Gujarati text & generate output file with stem words for each word in input file.
- Stop words which don't have much contribution in any language processing or information retrieval task, all removed in preprocessing & then stemmer algorithm is applied on remaining words.