The intended audience is practioners, researchers and graduate students working on clustering who wish to understand the new developing area of incorporating constraints into clustering. Knowledge of some popular clustering algorithms, e.g., KMeans, would be useful but is not essential - the tutorial will explain the algorithms used for illustration of the core ideas.
Since we believe a KDD audience will contain more practioners than an ICDM audience we will incorporate more real world applications, discussed in the literature, as case studies.
We believe there will be significant interest in the tutorial topic. Semi-supervised learning has received considerable attention in both the data mining and machine learning community. Davidson co-organized a special track at IEEE International Tools and A.I. (ICTAI) and was on the PC of the recent ICML 2005 workshop on the topic. Basu has presented at a similar ICML 2004 workshop. In addition the best paper awards at ICDM 2004, KDD 2004 and SIAM DM 2005 were all won by clustering under constraints papers.