Publishing open statistical data

Abstract

Publishing open data is going mainstream. There are diverse initiatives, ranging from international agencies to local governments, exposing data publicly in standard formats to foster transparency, innovation and public scrutiny. Nevertheless, the publishing of statistical data still presents huge challenges. Statistical modeling, privacy concerns, statistical secrecy, business models, etc., are still barriers to publishing massive amounts of statistical data without a corresponding requirement for significantly large processing. To date, access is usually restricted to limited views of the data. This paper presents a proposal for publishing census data, by using (1) the W3C open standard RDF; and (2) the design principles of Linked Data. The advantages of such an approach include the flexibility of the data which allows any user to adapt it to diverse applications for disparate purposes; the discovery of new relationships among the data; the accessibility of pieces and views of the data; the possibility of integrating it with other data; and last but not least, the possibility of providing simple services for complex querying, demos and visualizations.