jixtes/Amharic-rule-based-lemmatizer

# Amharic-rule-based-lemmatizer Stemming is the process of reducing affixes to the stem of the word that can serve different inflictions and sometimes derivation the word can possibly go through, producing different morph variations. Lemmatization is reduction upto a dictionary form. This program works on python3. The lemmatizer consists of 3 distinct stages; Transliteration, Stemming and Disambiguation. The unicode input first gets transliterated and then goes through a series of predefined rule checks for affixes and stemming, after the stemming we may have maximum of 10 and minimum 1 stem candidates. These candidates will then be searched in the MRD(machine readable dictionary) in­order to disambiguate and choose the right stem. The longest string length matched will be chosen automatically. This occurrence data is important in the disambiguation process of the stemmer when two matches are found in the dictionary with the same string length.