Rules play a fundamental role in computer science; for example, knowledge-based systems are based on rules. Usually, rules are considered as static elements, without any temporal component. In this talk I will approach the problem of extracting temporal rules from temporal data. I will briefly recall the ideas behind the algorithm APRIORI, and, then, present the following issues: 1) how to model temporal data towards temporal rule extraction? 2) how to represent temporal rules? 3) how to adapt the algorithm APRIORI to this end? I will present a solution to these issue, with examples of applications.
Guido Sciavicco holds a MSD and a PhD in Computer Science from the University of Udine (Italy). He has been working as a Junior and Senior research fellow at the University of Murcia, as Visiting Professor at the UIST of Ohrid (Macedonia) and at the METU of Guzelyurt (Cyprus), and he is now Associate Professor at the University of Ferrara (Italy). He is interested in foundational aspects of theoretical computer science, especially in logic-based aspects. His main line of research has been the study of the computational behaviour of several temporal algebras and logics; recently, he has been interested also in machine learning, and, in particular, in enhancing classical machine learning algorithm to deal with temporal aspects of data. He co-authored about a hundred publications in theoretical computer science, logic, artificial intelligence, and intelligent data analysis in the past 15 years.