Duplicate Records Detection Techniques: Issues and Illustration
18 Pages Posted: 16 Aug 2011
Date Written: August 16, 2011
Data quality has become a key issue in computer-based management systems. Inadequate data causes serious operational difficulties as well as direct financial losses. Furthermore, databases are increasing in size at an exponential rate with the proliferation of business electronization (Vasarhelyi and Greenstein 2003), globalization, and real-time economy. Once-manual processes are now stored in electronic databases, with data from multiple sources. This heterogeneity gives rise to a new set of problems. Errors – such as incorrect data entry, incomplete information, and unstandardized formats from different data sources – can lead to the existence of more than one representation of the same real-life object in a database. This paper discusses a problem that has received little academic attention but remains of great importance to an organization’s data quality, in particular, duplicate payments. We use data from a telecommunication company to demonstrate record matching techniques.
Suggested Citation: Suggested Citation