The Effects and Interactions of Data Quality and Problem Complexity on Data Mining

Data quality remains a persistent problem in practice and a challenge for research. In this study we focus on four of the most important dimensions of data quality accuracy, completeness, consistency, and timeliness. Definitions and conceptual models for these dimensions have not been collectively considered with respect to data mining in general and a key determinant of data mining outcomes, problem complexity, in particular. Conversely, these four dimensions of data quality have only been indirectly addressed by data mining research. Using definitions and constructs of data quality dimensions, our research shows for the first time that data quality and problem complexity have significant interaction effects on data mining outcomes. It also shows that the dimension of consistency can be concluded to have the largest effects of all four. Our results help address an important research challenge noted by March and Smith [15]. They also enable practitioners to assess data quality, prioritize efforts to improve it, and increase the performance of data mining for better decisions.