Digital revolution produces massive, heterogeneous and isolated data. These latter remain underutilized, unsuitable for integrated querying and knowledge discovering. Hence the importance of this survey on data integration which identifies challenging issues and trends. First, an overview of the different generations and basics of data integration is given. Then, semantic data integration is focused, since it semantically links data allowing wider insights and decision-making. More than thirty works are reviewed. The goal is to help analysts to identify relevant criteria to compare then choose among semantic data integration approaches, focusing on the category (materialized, virtual or hybrid) and querying techniques.

数字革命产生海量、异构、孤立的数据。后者仍未得到充分利用，不适合集成查询和知识发现。因此，这项关于数据集成的调查非常重要，它确定了具有挑战性的问题和趋势。首先，概述了数据集成的不同世代和基础知识。然后，语义数据集成成为焦点，因为它在语义上链接数据，从而允许更广泛的见解和决策。审查了三十多部作品。目标是帮助分析师确定相关标准进行比较，然后在语义数据集成方法中进行选择，重点关注类别（物化、虚拟或混合）和查询技术。