Assuring Data Quality by Placing the User in the Loop

Advanced analytical techniques such as data mining, text mining or predictive analytics are concepts that are increasingly important in the area of discovering large data sets. Various business areas recognize that data in all formats and sizes can provide significant support for decision-making. Large amounts of data can contain explicit knowledge in form of patterns. Errors within the data can falsify extracted patterns. Data is useful if it is correct, organized and interpreted correctly. Data mining algorithms can help improve data quality. Algorithms can suggest hints on possible errors. Possible errors need a mechanism that decides whether the error is true or false. The solution this paper introduces is to integrate users in the quality assurance process for decision support systems. The user can assess whether an error is true or false.