High-performing feature selection for text classification
Proceedings of the eleventh international conference on Information and knowledge management
A Tutorial on Support Vector Machines for Pattern Recognition
Data Mining and Knowledge Discovery
An extensive empirical study of feature selection metrics for text classification
The Journal of Machine Learning Research
Feature selection and feature extraction for text categorization
HLT '91 Proceedings of the workshop on Speech and Natural Language
Entity discovery and assignment for opinion mining applications
Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining
| Hi-index | 0.00 |
In this paper, Harmony Search is introduced as a meta-heuristic that has a stronger presence with respect to intensification and a tuned version of Genetic algorithms, with respect to diversification. An approach to solving the Feature selection (FS) problem using Harmony Search is proposed. The problem of Dominant Feature Selection (DFS) is introduced with respect to product reviews and two solutions to the problem, one based on Genetic Algorithms and another, based on Harmony Search are proposed. By experimental evaluation, we conclude that the Feature Selection problem is best solved by a meta-heuristic that is stronger with respect to intensification (Harmony Search in our case) and that the Dominant Feature Selection is best solved by a meta-heuristic that is stronger with respect to diversification. This paper aims to give a brief guide to the judicious choice of meta-heuristics to solve problems in Text Mining.