A Hybrid Approach to Text Categorization Applied to Semantic Annotation

摘要

In this paper, a hybrid approach is presented as a new technique for text categorization, based on machine learning techniques such as Vector-Space Model combined with n-grams. Given a specified content this technique takes care of choosing from different categories the one that best matches it. FLERSA is an annotation tool for web content where this technique is being used. The hybrid approach provides to FLERSA the capability for automatically define semantic annotations, determining the concepts that the content of a web document deals with.