A Multi-View Clustering Trust Inference Approach Using Gray Affinity Model

In recent years, Multi-view Affinity Propagation (MAP) methods are important and widely accepted techniques which measure the within-view clustering and clustering consistency across different view. However, these systems suffer from several inherent shortcomings such as similarity and correlation between clusters. With the development of recommender systems, trust and similarity measured introduced as a new approach to overcome the problem. But these approaches suffer from relatively low accuracy and especially coverage too due to avoidance of implicit trust. Therefore to address these problems, in this paper we propose a framework called, Multi-View Clustering based on GrayAffinity (MVC-GA)by integrating both similarity and implicit trust. Firstly, similarity between two clusters is obtained by applying Pearson Correlation Coefficient-based Similarity. Then, it utilizes the Collaborative Filter-based Trust evaluation for each clustered view in terms of the similarity based on Gray Affinity NN algorithm. Classification of incomplete occurrences is addressed based on Gray Affinity Function. Experiments on the benchmark data sets have been performed to validate the proposed framework. The experimental results on this data sets show that MVC-GA can effectively improve both the multi-view clustering accuracy and coverage. The promising results demonstrate the effectiveness of our framework.