Authors: S. Rajini; A. Vasuki
Addresses: Kumaraguru College of Technology, Coimbatore, Tamil Nadu, 641049, India ' Kumaraguru College of Technology, Coimbatore, Tamil Nadu, 641049, India
Abstract: In the field of computational linguistics, word sense disambiguation (WSD) is a problem of high significance which helps us to find the correct sense of a word or a sequence of words based on the given context. It is treated as a combinatorial optimisation algorithm wherein the aim is to discover the set of senses which help to improve the semantic relatedness among the target words. Nature inspired algorithms are helpful to find optimal solutions in reduced time. They make use of collection of agents that interact with the surrounding environment in a coordinated manner. In this article, two such algorithms, namely, cuckoo search and firefly algorithms, have been used for solving this problem and their performance have been compared with the D-bees algorithm based on bee colony optimisation algorithm. They have been evaluated using the standard SemEval 2016 task 11 dataset for complex word identification. Experimental results show that firefly algorithm is performing better than the other algorithms.
Keywords: word sense disambiguation; WSD; cuckoo search; optimisation; firefly; bees algorithm; unsupervised.
International Journal of Cloud Computing, 2021 Vol.10 No.1/2, pp.78 - 89
Received: 14 Jun 2019
Accepted: 23 Dec 2019
Published online: 15 Mar 2021 *