Data Source Selection Support in the Big Data Integration Process - Towards a Taxonomy

Selecting data sources is a crucial step in providing a useful information base to support decision-makers. While any data source can represent a potential added value in decision making, it’s integration always implies a representative effort. For decision-makers, data sources must contain relevant information in an appropriate scope. The data scientist must assess whether the integration of the data sources is technically possible and how much effort is required. Therefore, a taxonomy was developed to identify the relevant data sources for the decision-maker and minimize the data integration effort. The taxonomy was developed and evaluated with real data sources and six companies from different industries. The final taxonomy consists of sixteen dimensions that support the data scientist and decision-maker in selecting data sources for the big data integration process. An efficient and effective big data integration process can be carried out with a minimum of data sources to be integrated.