Conflict Identification and Resolution in Heterogeneous Datasets: A Comprehensive Survey

Abstract

Data Integration has become the vital necessities of today's interconnected world. Information is scattered everywhere and to retain the strategic advantage, it becomes mandatory for organizations to obtain as much information as possible. Hence combining the scattered data sources to obtain information becomes the only solution. Data integration is posed by several challenges including the basic nature (heterogeneity) of the data. This paper describes the basic elements of a data integration system and emphasizes on the data fusion phase which forms the core functionality of the architecture. The problems occurring during data fusion (conflicts) are discussed and it also provides a comprehensive survey of the techniques used to resolve conflicts. Functionalities lacking in the current system and future research directions are discussed in detail.