A data integration system provides a uniform interface to distributed and heterogeneous sources. These
sources can be databases as well as unstructured information such as files, HTML pages, etc. One of the most
important problems within data integration is the semantic heterogeneity, which analyzes the meaning of
terms included in the different information sources. This survey describes seven systems and three proposals
for ontology -based data integration. An important feature is that all of them use, in some way, ontologies as
the way to solve problems about semantic heterogeneity. In this paper, we show similarities and differences
among the systems by providing a framework for comparison and classification.