Continuing this year’s theme of ‘openness’, the second NetIKX seminar of 2019 was on the topic of Open Data and was presented by committee member David Penfold. David began by giving a definition of open data and clarifying what open data is not (it is often confused with open access). Using the DIKW (data, information, knowledge, wisdom) model, he explained that data is the foundation of this model: without the underlying data, the rest is meaningless. This was followed by a quick tour of the history of data classification and organisation, from Aristotle’s Tree of Porphyry through to Tim Berners-Lee’s concept of the Semantic Web. It was interesting to see the forerunners of what we often consider to be relatively recent ideas, such as Peirce’s ‘existential graphs’ which bear a striking similarity to RDF triples!
The presentation included two case studies, one from Network Rail and one from Parish Online, both focusing on how the use of (linked) open datasets can improve access to information for both end users and organisations by breaking down silos. This led into a discussion of the value of linked data and the Semantic Web, as well as considerations around the ethical use of data, particularly in the context of AI. The Open Data Institute has developed a Data Ethics Canvasto help organisations identify such issues.
In the syndicate sessions we were invited to explore questions such as whether our organisations use or provide open data (and why/why not?), what aspects of the Data Ethics Canvas we considered most important, and whether AI has changed, or will change, the way we handle data. This resulted in a wide-ranging discussion in which we also considered issues such as who owns and controls ‘our’ data and the ways in which geolocation can be used for both positive and negative purposes. Many thanks to David for single-handedly leading a seminar on these complex and sometimes challenging issues, and to all those who contributed to the discussion.
A collection of the tweets from this seminar can be found on Wakelet.