Multi-Label Text Categorization Using a Probabilistic Neural Network

Abstract

Techniques for categorization and clustering, range from support vector machines, neural networks to Bayesian inference and algebraic methods. The k-Nearest Neighbor Algorithm (kNN) is a popular example of the latter class of these algorithms. Recently, slightly modified versions of support vector machines, kNN and decision trees have been proposed to deal better with multi-label classification problems. In this paper, we also proposed a new version of a Probabilistic Neural Network (PNN) to tackle these kind of problems. This PNN was proposed aiming at executing automatic classification of economic activities, which is the focus of this article. Nevertheless, we compared the PNN algorithm against other classifiers. In addition to economic activities database, we applied our algorithm to some other databases found in the literature. In general, our approach surpassed the other algorithms in many metrics typically well known in the literature for the multi-label categorization problems.