Overcoming database heterogeneity to facilitate social networks: the Colombian displaced population as a case study

Abstract

In this paper we describe a two-step approach for the publication of data about displaced people in Colombia, whose lack of homogeneity represents a major barrier for the application of adequate policies. This data is available in heterogeneous data sources, mainly relational, and is not connected to social networking sites. Our approach consists in a first step where ontologies are automatically derived from existing relational databases, exploiting the semantics underlying the SQL-DDL schema description, and a second step where these ontologies are aligned with existing ontologies (FOAF in our example), facilitating a better integration of data coming from multiple sources.