Surrey County Council

Surrey Linked Open Data portal

Briefly describe the initiative/ project/ service

Surrey’s Linked Open Data portal is designed to host comprehensive and curated data for anyone to access, use and share. It conforms with best practise in Open Data publication, as defined by Sir Tim Berners-Lee’s 5* ranking scheme and semantic web principles. Part of Surrey’s Digital Transformation, the project aimed to test new ways of working, challenge attitudes towards data sharing, build skills in data manipulation, observe common web standards and to generally improve the access to and curation of our information. The portal hosts data in both 3* and 5* formats, suiting casual and more advanced audiences. 5* linked open data conforms with W3C standards and uses recognised global definitions, vocabulary sets and ontologies. This includes links to complimentary information (e.g. DCLG, ESD) via URIs and use of RDF triples to model relationships in a way that machines can meaningfully understand. Our published data forms part of the semantic web where it can be automatically consumed by computers rather than necessitating human intervention. The Linked nature means that it is possible to conduct a federated query to retrieve information from across the wider linked data web, thereby significantly embellishing our own data.

Why do you think it should win this award?

The portal was the first of its kind to publish all Transparency Code data at the top-most level (5*). Many datasets were also published with more than the mandatory level of detail, for example, all spend items over £250. Surrey was nominated as ‘Open Data Champion’ in 2015 by the then Cabinet Office Minister, Francis Maude. Publication of data at 5* level involved considerable attention to data standards and modelling. This included recording comprehensive metadata (also machine readable) about each dataset and considering relationships to other datasets and organisations, within the sector, country and worldwide. Datasets are typically licensed according to the Open Government Licence, thereby granting favourable access to end-users, developers, startups, academics and civic entrepreneurs. During 2 months in 2015, maintenance of the portal was performed by a work experience placement from ‘Surrey Choices’ with learning difficulties. The relational data modelling, and trust given by our team, was of particular interest/benefit to this student and they left with a sense of fulfilment from having worked on something that was high value to the authority.

What are the key achievements?

Discovery of potential datasets to publish lead us to many high public interest areas (e.g. highways, schools). Being able to show the benefits of the agile project in its alpha form meant that we could reassure colleagues of how data would be transported, presented, licensed and what value that would provide end-users. Datasets on the portal can be queried via SPARQL API, including the ability to view RDF data in many different serialisations. Thanks to machine-readability and use of web standards, we have been able to setup automated data harvesting to, including 5* RDF and 3* CSV datasets. We established relations with ODI startup ‘SpendNetwork’ on the basis that mutual gains could be achieved from improving the quality and depth of our spend data. This involved us improving the standardisation of our data, for example by publishing ProClass cross ref codes. It also focussed our data teams on the importance of good data curation as well as emergent technology and techniques (i.e. decision-tree machine learning). We have written complete documentation of our open data procedures which details local conventions. These have been shared with the LGA, ODI, DCLG and Councils, contributing to the LGA’s web learning modules