Centrality Measures for Text Clustering

Abstract

Text clustering is an unsupervised process of classifying texts and words into different groups. In literature, many algorithms use a bag of words model to represent texts and classify contents. The bag of words model assumes that word order has no signicance. The aim of this article is to propose a new method of text clustering, considering links between terms and documents. We use centrality measures to assess word/text importance in a corpus and to sequentially classify documents.

Keywords:

• Centrality measures
• Term weighting models
• Text clustering