Multiple clustering solutions. A large body of work studies the problem of discovering multiple clustering solutions [26], [27], [28], [29], [30]. The objective in these papers is to discover multiple clusterings for a given dataset. Each of the clusterings needs to be of high quality and the clusterings are required to be different with each other in order to cover different aspects of the dataset. Each of the discovered clusterings is non-overlapping, so this clustering paradigm is not directly comparable with our clustering result.