Semantic classification of attributes for integrating heterogeneous relational databases with artificial neural networks

Abstract

One of the major problems in heterogeneous databases integration with different scheme and management system, is the data tautology, which involves an issue of semantics in the content mainly from the names and addresses at the time of normalizing the database, this problem is tackled in this article, by means of a neural network to classify the data. The results show that can be proper classification of up to 84% with these results may normalize the database and properly solve the problem generated by the database integration..