Association discovery is one of the most studied tasks in the field of data mining (Agrawal et al. 1993).
It involves identifying items that occur together in data, and has numerous applications in manufacturing, commerce, administration and science. However, far more attention has been paid to how to discover associations than to what associations should be discovered. In this talk Geoff will provide a highly subjective tour of the field. He will
• highlight shortcomings of the dominant frequent pattern paradigm;
• illustrate benefits of the alternative top-k paradigm; and
• present the self-sufficient itemsets approach to identifying potentially interesting associations as described in Webb & Zhang (2005), Webb (2007, 2008, 2010, 2011).