Title: Text document classification using a hybrid approach of ACOGA for feature selection
Authors: Avjeet Singh; Anoj Kumar
Addresses: Department of Computer Science and Engineering, MNNIT Allahabad, U.P., India ' Department of Computer Science and Engineering, MNNIT Allahabad, U.P., India
Abstract: Categorisation of text document involves a great deal of information or features. Feature selection is normally used to reduce the dimensionality of datasets that have a large number of attributes or features. Appropriately, feature determination is vehemently considered as one of the pivotal parts of text document classification. There are many methodologies employed by various researchers to handle this issue. In this paper, we present an approach for categorisation of text documents using a k-NN classifier that works upon the best-selected features of the document. The feature selection process is advanced that lessens the dimensionality of the feature space and thus increment the performance. The feature selection is done through a hybrid approach that utilises the combination of ant colony optimisation (ACO) algorithm and genetic algorithms (GA). The feature selection process is optimised that reduces the dimensionality of the feature space and consequently increase the performance.
Keywords: text categorisation; ant colony optimisation; ACO; genetic algorithm; GA; feature selection.
DOI: 10.1504/IJAIP.2021.117613
International Journal of Advanced Intelligence Paradigms, 2021 Vol.20 No.1/2, pp.158 - 170
Received: 19 Jun 2018
Accepted: 15 Jul 2018
Published online: 16 Sep 2021 *