Generating Linked Data from Tables.

Large amounts of information is stored in tables, spreadsheets, CSV files and databases for a number of domains, including the Web, healthcare, e-science and public policy. The tables' structure facilitates human understanding, yet this very structure makes it difficult for machine understanding. This talk will focus on describing our work on making the intended meaning of tabular data explicit by representing it as RDF linked data, potentially making large amounts of scientific and medical data in important application domains understandable by machines, improving search, interoperability and integration. The talk will briefly cover our domain-independent framework which uses background knowledge from the LOD cloud to jointly infer the semantics of column headers, table cell values (e.g., strings and numbers) and relations between column and represent the inferred meaning in RDF. Specifically the talk will focus on a probabilistic graphical model which is at the core of this framework. It will also cover our proposed improvement on a message passing scheme that exploits semantics during the joint inference process. The talk will also present some preliminary evaluations on this new message passing scheme and present future directions.