Adapting Sequence Alignments for Text Classification

Abstract

Text classification is still an area growing as text sizes grow rapidly with the information and internet revolution. The majority of conventional text classification approaches are hierarchical training, where the classification system explicitly differentiates between groups. However, this is different from the nature of language processing which suffers from ambiguity and lack of clarity. That is, the instance could be of more than one class. This paper handles the problem of text classification by applying a novel classification method based on sequence alignment with simple fuzzy concepts. The experiments showed expected performance compared to other conventional classifications of natural languages.