Discovering periodic and correlated high utility patterns in customer transaction databases Event as iCalendar

(Science Event Tags, Seminars, Computer Science)

04 August 2017

11am - 12pm

Venue: 303s.561

Speaker: Professor Philippe Fournier-Viger 


A popular data mining task is high utility itemset mining. It consists of discovering sets of items (products) purchased together that yield a high profit in customer transaction databases. Although many algorithms have been published for identifying high utility itemsets in transactions, many of those algorithms have important limitations such as not considering the time dimension, and finding itemsets containing items that are weakly correlated.

In this talk, Professor Fournier-Viger will discuss extensions of the high utility itemset mining problems that are designed to discover more meaningful patterns. He will first present the problem of periodic high utility pattern mining, which aims to discover recurring customer behaviour that yields a high profit (e.g. a customer buys some products every week). He will also present an algorithm to discover correlated high utility itemsets (items that yield a high profit and are strongly correlated). Finally, he will briefly mention the problem of considering the shelf time of products to discover itemsets that are profitable during specific time periods.


Philippe Fournier-Viger is a Canadian researcher, full professor at the Harbin Institute of Technology in Shenzhen, China, and adjunct professor at University of Moncton, Canada. His research interests include data mining, frequent pattern mining, sequence analysis and prediction, big data and applications. He has received the title of 'Youth 1000 talent' from the National Science Fundation of China. He has published more than 150 research papers in refereed international conferences and journals, which have received 1000 citations in the last three years. He is the founder of the popular SPMF open-source data mining library (, which has been used in 450 research papers since 2010. He is also editor-in-chief of the Data Mining and Pattern Recognition journal.