Finding Alternative Clusterings Using Constraints

Improving Alternative Text Clustering Quality in the Avoiding Bias Task with Spectral and Flat Partition Algorithms

“…In [15], Davidson and Qi present an approach to finding alternative clustering which also uses constraints, in this case to characterise the grouping to be avoided. A distance function matrix is learnt from these constraints, which is decomposed afterwards using Singular Value Decomposition (SVD).…”

“…If we are trying to avoid the tendency (bias) of the clustering algorithm to fall in a certain grouping of the data that is being clustered the task is called Avoiding Bias. This problem has been tackled by several authors in the last years, which have proposed a wide range of approaches, ranging from distance learning [2] to using constraints [3]. However, it should be underlined that avoiding bias is still a clustering process, where the main focus is providing the user with a meaningful grouping of the data.…”

Avoiding Bias in Text Clustering Using Constrained K-means and May-Not-Links

“…Some papers approach the inclusion of the constraints through the learning of distance functions [16], such as Davidson and Qi [11], which uses Must-Link and Cannot-Link knowledge but implies the use of Singular Value Decomposition (SVD), or Cui et al [12], an approach to produce multiple orthogonal clustering views using Principal Component Analysis (PCA).…”

“…This has been a very fruitful field in the last years [4][5][6][7][8][9][10][11][12]. This constrained clustering is quite different from a classification process, as the domain knowledge gives the clustering algorithm rules over data instances (documents), instead of examples of the categories.…”

MultiClust special issue on discovering, summarizing and using multiple clusterings

“…Discovering multiple clusterings is an emerging topic that has received significant attention in recent years at top international conferences. The seminal paper on the topic won the best paper award at ICDM 2004 (Gondek and Hofmann 2004) with follow-up work establishing various research directions in meta-clustering (Caruana et al 2006), alternative clustering (Bae and Bailey 2006), alternative clusterings using constraints (Davidson and Qi 2008), multi-view clustering via orthogonalization (Cui et al 2007), non-redundant subspace clustering (Assent et al 2008), disparate clustering (Hossain et al 2010), and non-redundant spectral clustering views (Niu et al 2010). All of these research directions on multiple clusterings provide a new way of looking at the clustering problem.…”