Lifting Tabular Data to RDF: A Survey

Abstract

Tabular data formats (e.g. CSV and spreadsheets) combine ease of use, versatility and compatibility with information management systems. Despite their numerous advantages, these formats typically rely on column headers and out-of-band agreement to convey semantics. There is clearly a large gap with respect to the Semantic Web, which uses RDF as a graph-based data model, while relying on ontologies for well-defined semantics. Several systems have been developed to close this gap, supporting the conversion of tabular data to RDF. This study is a survey of these systems, which have been analyzed and compared. We identified commonalities and differences among them, discussed different approaches and derived useful insights on the task.