An Analysis of Data Integration Challenges from Heterogeneous Databases

Abstract

The Internet generates very large amounts of structured and unstructured data which creates storage, maintenance, management, sharing, privacy and security challenges. In the world today, organizations exchange and merge different types of data at a centralized location for the purpose of analysis and benefitting the organisations and individuals in achieving their business, economic, social, educational, cultural, and health objectives. The data merging or integration is a challenging process because of different type of data formats, structures, models, schemas, entities, attributes, and features. Integration is a complex and tedious process, and involves a number of technologies and extensive processing, and so it is not straightforward to integrate very large data with a variety of data formats and types. This paper discusses issues and complexities faced in data integration processes. The paper also discusses different methods of data integration, their advantages and disadvantages, and provides a comparative analysis to gain better insights from examples of recently completed projects.