Mining Multirelation Association Rules on the Web of Data

• Main Author: de Oliveira, Felipe Alves
• Publication Date: 2020
• Other Authors: Villote, Guilherme dos Santos, Costa, Raquel Lopes, Goldschmidt, Ronaldo Ribeiro, Cavalcanti, Maria Cláudia
• Format: Article
• Language: por
• Source: Brazilian Journal of Information Systems
• Download full: https://journals-sol.sbc.org.br/index.php/isys/article/view/830
• topic: Multirelation Association Rule Mining; Web of Data; Graph Mining; Data Mining; Mineração de regras de associação de multirrelação; Web de Dados; Mineração de Grafos; Mineração de Dados

The Web of Data is a growing and relevant data source that contains information distributed in interconnected datasets. Most data mining algorithms were designed to analyze a single dataset at a time and, hence, cannot explore the connections between the datasets in the Web of Data. To overcome this limitation, the present work proposes MRAR+, a graph mining method that searches for multirelation association rules in order to identify new and useful knowledge involving resources of multiple datasets connected to the Web of Data. To illustrate MRAR+'s feasibility, this paper reports on two experiments where the proposed method mined different datasets interconnected in the Web of Data and produced new and useful rules for the users.