What should an open data platform look like? This was the question addressed by a session triggered by Martin Howitt of ODI Devon in relation to Bristol’s plans to develop its platform.
“[There is] an open data platform that was brought in by the council [to release] data sets, but it wants to move on to next level: e.g. citizen sensing projects, citizen data sets, information from third sector and even private sector partners,” he said.
“At the same time, there’s a move towards creating a more general data infrastructure. So there’s a new control centre going in, streaming from sensors, wild and whacky stuff coming in from different places. So the question is: what requirements does this lead to?”
Practical & technical
Kev Kirkland from Data Unity stressed that the council really wanted practical examples of things to include. Ideas, once they started to flow, ranged from deeply technical requirements to ways to help citizens find and use data relevant to them.
On the technical front, one participant said: “I am involved with portal for Ireland [and it] puts up information about licences, which says ‘this is the data and this is what you can do with it…’ He argued that this would make it easier for people to handle data with confidence.
Other participants suggested that data management tools should be built into the platform, to provide “hooks” to pull out information items that related to an issue of interest that were included in different data sets, uploaded by different providers at different times, along with effective and multiple tagging.
There was also enthusiasm for forums and feedback, to enable developers to ask questions about the data; and the providers to respond and provide updates. Martin asked whether a register of who had used APIs were a good idea.
Opinions were divided, with some session members feeling that having to register and get an API key would act as a barrier for those who wanted to use the data, and others arguing that it could encourage dialogue and prove the benefit of the whole venture.
Demystifying the data
On the citizen front, ideas included: “demystifying the data” by explaining basic terms, such as ‘CSV’; adding contextual information to the data sets, such as newspaper articles, that would explain why the data might matter; providing signposts, such as colour-coding for data sets like those used in the Datapolis game; and “providing a really big search box that works really well.”
Kev asked whether any sites already pulled off this trick “that we could just go and nick.” Suggestions included the New York parking fines site, the London Data Store, and the NHS’ attempts to open up APIs that used its data.
But, overall, one session member pointed out that there were really a number of users whose needs had to be considered. “You need something really technical and you need something that looks nice. You don’t want something that’s so tough only a few people use it, but you don’t want it to look really nice but to have nothing behind it.”