Granular Computing for Text Mining: New Research Challenges and Opportunities

As an emerging computational methodology, granular computing provides an effective strategy for solving many real world problems such as mining latent relationships from text. This paper examines the relationship between granular computing and text mining from a theoretical perspective. Firstly, we analyzes the granular structure of text data which is the key step for textual data representation. Secondly, some granule-based computational methods are described, especially on term-document and document-document similarity calculations. Finally, we highlight several potential research areas where the performance of text mining could be enhanced by applying the concepts of granular computing.