Post-Processing Association Rules using networks and label propagation

• Grant number: 13/12392-0
• Support Opportunities: Scholarships in Brazil - Master
• Start date: October 01, 2013
• End date: February 28, 2015
• Field of knowledge: Physical Sciences and Mathematics - Computer Science
• Principal Investigator: Solange Oliveira Rezende
• Grantee: Renan de Padua
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

In recent years, data mining has emerged as a powerful tool in extracting interesting knowledge that is "hidden" in the data. Among the existing techniques for data mining is association, association is responsible for identifying relationships that occur in the data set. Although the association is one of the most widely used techniques, the amount of extracted patterns may ultimately overwhelm the user's ability of understanding so that the exploration process becomes infeasible. To solve this problem, several approaches have been proposed in the literature, and the most recents include the user in the process, the user has the opportunity to represent their prior knowledge of the domain and, based on this knowledge, the post-processing is directed to the association rules considered potentially more useful. However, sometimes the user doesn't have the knowledge necessary to make such specifications, he might not know which relationships can be potentially interesting, which items he has specific interest, etc.. Furthermore, this specification makes the user lose a long time and does not guarantee that the domain is completely and correctly modeled. In this context, this research project proposes an interactive approach to post-processing of association rules based on the user knowledge and / or interest, to be provided implicitly during the process of exploration. The principle adopted is that such information wil be collected during the post-processing rules, through choices made by the user in order to direct the operation according to specific interests. In order to make the proposed approach, networks are used to model the rules and allow the spread of knowledge and / or interest to the user through its exploration properties. Finally, we hope to obtain as a result of this project an interactive approach to post-processing of association rules to make the identification of relevant standards to the domain easier.

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

PADUA, Renan de. Post-processing association rules using networks and label propagation. 2015. Master's Dissertation - Universidade de São Paulo (USP). Instituto de Ciências Matemáticas e de Computação (ICMC/SB) São Carlos.