Most of the clustering algorithms try to find a single efficient clustering solution. However, data can be grouped and interpreted in many different ways. Moreover, different efficient clustering solutions are interesting for different views. This is particularly true in the high dimensional, where different views reveal different possible groupings of the data. In this paper, we apply multivariate mutual information to find multiple views for the data by grouping features. Then we apply the Fast K-means (FK-means) algorithm to these different views to find multiple clustering solutions. Our experiments on synthetic and real data show that our method can discover multiple views of the data that correspond to different solutions