The evolution of heterogeneous data residing in various kinds of data sources (e.g., relational, XML, document stores, etc.) raises the need to provide a unified view over them. Ontology Based Data Access (OBDA) is a semantic web technology that comprises a set of algorithms and techniques for dealing with data heterogeneity. Ontologies are utilized in OBDA to provide a global view over various local datasets; and mappings are typically employed to define the relationships between them in two ways: data translation, query translation. In data translation, original data is transformed and materialized according to the global view. Whereas in query translation, queries written according to the global schema are transformed into the query language supported by the original data sources and evaluated in the original data management systems. In this paper, we present the framework of OBDA by discussing the main components of ontology-based data access, techniques, applications and future challenges.