MapReduce vs Bancos de Dados Paralelos no cálculo de medidas de centralidade em grafos

• Fabiano Fernandes FACCAMP
• Eduardo Yero FACCAMP

Abstract

In recent years we have witnessed an unprecedented explosion on the amount of data digitally available. A significant part of this data can be modeled using graphs, particularly data related to social networks. This article compares the effectiveness of two paradigms: the MapReduce model and the Parallel System Database when calculating eccentricity measures (radius and diameter) for a graph. Preliminary results indicate that Parallel System Databases are better suited to solve this problem than MapReducebased solutions.

References

Balouek, D., Lèbre, A., & Quesnel, F. (2013). Flauncher and DVMS -- Deploying and Scheduling Thousands of Virtual Machines on Hundreds of Nodes Distributed Geographically. Finalist of the IEEE International Scalable Computing Challenge (SCALE 2013).

Dean, J., & Ghemawat, S. (2008). MapReduce: Simplied Data Processing on Large Clusters. ACM Digital Library, 107-113.

Del Vecchio, R., Lima, L., Galvão, D., & Loures, R. (2009). Medidas de Centralidade da Teoria dos Grafos aplicada a Fundos de Ações no Brasil. XLI Simpósio Brasileiro de Pesquisa Operacional, Porto Seguro.

Nascimento, J. P., & Murta, C. (2012). Um algoritmo paralelo em Hadoop para Cálculo de Centralidade em Grafos Grandes. XXX Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos. Ouro Preto: SBC.

Stonebraker, M., Abadi, D., DeWitt, D., & Madden, S. (2010). MapReduce and parallel DBMSs: friends or foes? Communications of the ACM. 53(1):64-71.