Keynote talk

Tzung-pei Hong Chair Professor Tzung-Pei Hong
Department of Computer Science and Information Engineering
National University of Kaohsiung

Biography: Tzung-Pei Hong received his Ph.D. degree in computer science and information engineeringfrom National Chiao-Tung University in 1992. He served as the first director of the library andcomputer center, the Dean of Academic Affairs, and the Vice President in the National Universityof Kaohsiung, Taiwan. He is currently a chair and distinguished professor at the Department ofComputer Science and Information Engineering in NUK, and a joint professor at the Departmentof Computer Science and Engineering, National Sun Yat-sen University, Taiwan. He has published more than 600 research papers in international/national journals andconferences and has planned more than fifty information systems. He is also the board member ofmore than forty journals and the program committee member of more than one thousandconferences. His current research interests include knowledge engineering, data mining, softcomputing, management information systems, and www applications

Title: Duality of Mining Problems
Abstract: Frequent-itemset mining and erasable-itemset mining are two commonly seen and practicaltechniques for finding different useful itemsets in data mining. Frequent-itemset mining is asignificant pre-processing step in the search for association rules and is mainly conducted basedon the frequencies of the itemsets in a transaction database. On the other hand, erasable-itemsetmining is often applied to product production planning and identifies the itemsets that would notsignificantly impact production profits if removed. Although the two mining techniques seem tobe different and independent, we have derived they are actually equivalent to each other. Weformally prove that these two mining techniques possess the property of duality. We designmethods to transform one of the two mining problems into the other and then solve it, and viceversa. The mining results with and without the transformation will be identical. We then extendthe duality property of the two mining problems to multiple-threshold and weighted-itemsituations. With the duality property, we can easily design an algorithm from its correspondingone in the dual mining problem