Rough set theory for document clustering: A review

Rough set theory is a mathematical framework that can be visualized as a soft computing tool dealing with the vagueness and uncertainty of data and is applied to pattern recognition, data mining, and knowledge discovery. Document clustering is another area of research with values which are a bag of words that describe contents within clusters. This work analyzes how rough set theory is used for document clustering to fix issues that clustering methods manage. In this survey, an exhaustive literature review of the concept of rough sets, as well as how the lower and upper approximation of a set can be used for document clustering, has been presented. Rough set clusters are shown to be useful for representing real-time applications such as biomedical inferences, network data handling, and citation analysis. The survey is done in phases, showing how machine learning algorithms have been incorporated for document clustering using rough set theory, as well as how rough set theory has been extended to adapt to document clustering with feature selection techniques and feature/dimensionality reduction and, finally, ending with a view of assorted clustering tasks where rough set theory is applied. The classification of rough set theory for document clustering is depicted and its applications presented in this paper. The rough set theory works with resolving ambiguity and uncertainty in data. To the best of our knowledge, a rough set clustering survey has not been done earlier in the literature reviewed and the survey ends with a critical analysis of rough set theory in each application of clustering.