An Efficient Meta Heuristic Algorithm for POS-tagging

Tagging is an increasingly important task in natural language processing domains. As there are many natural language processing tasks which can be improved by applying disambiguation to the text, fast and high quality tagging algorithms are a crucial task in information retrieval and question answering. Tagging aims to assigning to each word of a text its correct tag according to the context in which the word is used. Part Of Speech (POS) tagging is a difficult problem by itself, since many words has a number of possible tags associated to it. In this paper we present a novel algorithm that deals with POS-tagging problem based on Harmony Search (HS) optimization method. This paper analyzes the relative advantages of HS metaheuristic approache to the well-known natural language processing problem of POS-tagging. In the experiments we conducted, we applied the proposed algorithm on linguistic corpora and compared the results obtained against other optimization methods such as genetic and simulated annealing algorithms. Experimental results reveal that the proposed algorithm provides more accurate results compared to the other algorithms.