Mineração de Regras de Associação Multirrelação em Datasets na Web de Dados

  • Felipe Alves de Oliveira Instituto Militar de Engenharia (IME)
  • Raquel Lopes Costa Instituto Nacional do Câncer (INCA)
  • Ronaldo R. Goldschmidt Instituto Militar de Engenharia (IME)
  • Maria Cláudia Cavalcanti Instituto Militar de Engenharia (IME)


The Web of Data is an extremely rich source that contains information of different types, distributed and organized in the form of graphs in interconnected datasets. Although several data mining algorithms have been developed to extract knowledge from large databases, typically such algorithms restrict themselves to analyzing data in a single dataset, which imposes a limitation on the possibilities of exploration of the various connections between the datasets in the Web of data in search of new knowledge. To overcome this limitation, the present work proposes the MRAR+, a method for mining multirelation association rules in graphs that, unlike the methods of the state of the art, is able to identify new and useful knowledge involving resources of multiple datasets connected to the Web of data. However, the proposed method seeks to identify resources linked to other datasets that, when considered in the mining process, have the potential to increase the chances of obtaining new and useful knowledge in the analyzed dataset. MRAR+ was implemented based on the MRAR algorithm that extracts multirelation association rules in graphs. In addition, MRAR+ was applied in two case studies and has produced new and useful rules for the user, illustrating the feasibility of the proposed method for mining different datasets interconnected in the Web of Data.
Palavras-chave: Multirelation Association Rule Mining, Web of Data
Como Citar

Selecione um Formato
DE OLIVEIRA, Felipe Alves; COSTA, Raquel Lopes; GOLDSCHMIDT, Ronaldo R.; CAVALCANTI, Maria Cláudia. Mineração de Regras de Associação Multirrelação em Datasets na Web de Dados. In: SIMPÓSIO BRASILEIRO DE SISTEMAS DE INFORMAÇÃO (SBSI), 15. , 2019, Aracajú. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2019 . p. 479-486.