Data mining techniques to study voting patterns in the US

Abstract

This paper presents data mining techniques that can be used to study voting patterns in the United States House of Representatives and shows how the results can be interpreted. We processed the raw data available at http://clerk.house.gov, performed t-weight calculations, an attribute relevance study, association rule mining, and decision tree analysis and present and interpret interesting results. WEKA and SQL Server 2005 were used for mining association rules and decision tree analysis.

Keywords

• Data mining
• Data preprocessing
• Attribute relevance study
• Association rule mining
• Decision tree analysis
• Voting patterns.