Abstract
Due to the size and complexity of today's database systems, identifying duplicate records in a data source is extremely important to many organizations. The data duplication problem has resulted in a significant amount of lost revenue in terms of disgruntled customers and incomplete sales orders, among others. A proposed solution to the data duplication problem is the use of neural network. In this paper, a back propagation or generalized delta rule is utilized to train a neural network to identify duplicate records in a given data source. Though this research is an ongoing effort, this approach offers great promise in resolving the data duplication problem.