Türkçe kelimelerin türlerinin belirlenmesi

Natural Language Processing (NLP) is a field of research that studies the problems of automated generation and understanding of natural languages. Part-of-speech (POS) tagging is a sub-problem in NLP which is interested in tagging the words in a document with the appropriate parts-of-speech. POS tagging has many uses in fields such as: full text searching, information retrieval, speech synthesis and pronunciation and high level text analysis. This study introduces TurPOS, a new rule-based part-of-speech tagger system that was developed for Turkish. The system aims to assign the appropriate word classes for each word in a given Turkish document. TurPOS uses a text corpora produced by a morphological analyzer as the input document. The system also uses a rule file that contains the list of grammatical Turkish rules. The structure of the rule file is quite simple and flexible. This makes it possible that the system can be used for tagging other languages, simply by modifying the rule file according to the grammar of the language.