Data Integration - Challenges, Techniques and Future Directions: A Comprehensive Study

Objectives: This paper studies various query reformulation techniques, which are used to convert the intermediate schema to the targeted schema. The techniques such as Ontology based information integration and data integration languages are also reviewed. Methods/Statistical Analysis: This paper discusses the techniques used for data integration and also to resolve inconsistencies from the integrated data. Data integration techniques mainly focusing on integration of data in several levels and applying independent or unified query over the data available. Findings: Analysis of various techniques done in the paper has led to the identification of several shortcomings and scope for improvements in the available techniques. This identified research directions includes vertical enhancement of wrappers by utilizing a single unified wrapper for all the data sources. Optimizing the queries depending on the data source is also another major requirement to provide efficient and faster results reducing the data retrieval latencies. The paper also advocates other research directions that include identifying duplicates from the retrieved data and performing effective elimination strategies to reduce space consumption. Identifying conflicts and applying strategies to eliminate conflicts is another major area with a huge scope for improvement. Application/Improvements: The comprehensive survey also recommends further works in the area of data integration techniques.