Co-similarity Approach to Co-clustering: Application to Text Mining and Bio-informatics

Clustering is an important tool for many applications such as document clustering, gene expression analysis, etc. In many such cases, the data can be represented as a set of instances expressed by their attributes, in the form of a matrix. Clustering instances, such as documents, depends on their attributes (words) and vice versa, thus forming a dual relationship between instances and their attributes. Co-Similarity Approach to Co-Clustering emphasizes on a technique that exploits the dual nature between instances and their attributes to find similarities between objects in each set. It provides and analyzes results of applying this technique on two different domains- Document clustering and Gene Expression Analysis.