A New Multimedia Documents Clustering Approach based on Feature Patterns Similarity

With the rapid advances in digital technology, the multimedia documents have been growing ubiquitously. The analysis of this huge repository of multimedia documents requires efficient organization of documents. Multimedia document clustering organizes the multimedia documents with common multimedia topics. The important step of multimedia document clustering is computing the similarity between multimedia documents. The multimodal objects of multimedia documents are described by the feature patterns. Hence, the feature patterns of multimedia objects play the major role in computing the similarity of multimedia documents. In this paper, we propose an feature pattern similarity based clustering approach for multimedia documents. Experimental results show that the proposed clustering approach clusters the multimedia documents efficiently and outperform the competitive methods.