Skip to Main Content
Building a generic Part-of-Speech (POS) tagger system without a lexicon (dictionary) depends on the language and the characteristics of its grammar, both the morphological and the syntactical systems of that language. Arabic language has a valuable and important feature, called diacritics, which are marks placed over and below the letters of Arabic word. This paper presents a novel algorithm to assign the correct POS tag to those words belonging to a verb or a noun class in an Arabic text. The algorithm is based on the pattern (wazn) of the word instead of using a huge manually tagged lexicon from which large amounts of training data can be extracted. An experiment was ran on a data set that contains 5,000 words belonging to a noun and a verb class to evaluate the accuracy of the algorithm. The algorithm is achieved an accuracy of 91%.