Automatic Stemming of Some Arabic Words Used in Persian Through Morphological Analysis Without a Dictionary

Abstract

Persian language is teeming with Arabic words and there is a need for Iranians to have access to some instrument which helps them differentiate between the Persian and foreign words. One such instrument is stemmer. A good stemmer for Persian must detect and stem these words properly. Such stemmers are by no means  free  from  problems.  The  basic  problem  for  stemming  these  words,  with  respect  to  Arabic,  is  their development  and  the  changes  they  go  through.  Morphologically,  Arabic  words  have  different  derivational behavior as compared with those of Persian. Furthermore, some of these words in Persian have specific features which help us distinguish them from Arabic words. To achieve the proper results, we have restricted ourselves to the derivation of some regular triliteral roots. The findings of this research can be utilized in the areas of information  retrieval,  text  categorization,  text  summarization,  automatic  detection  of  phrasal  categories, translation studies, natural language processing, etc.