Caracterizando desafios de interação com sistemas de mineração de regras de associação

Abstract

Data mining focuses on extracting useful information from great volumes of data, and thus has been the center of great attention in the recent years. Among the many techniques available for data mining, identifying association rules is one of the most popular. The novel aspect of rule association mining systems brings new challenges to the HCI field. In this article, we identify these challenges and analyze them based on the theory of action, and characterize it within the semiotic engineering theoretical framework. Thus, we provide designers with an explanation of aspects to be considered during use and design of such systems. This theoretical based explanation contributes to a deeper understanding of the issues involved in interacting with association rules mining systems, allowing for better informed decisions during design process. It also motivates future empirical investigations.