Day two of Open Data Camp in Cardiff opened with another session on the basics. What is open data, who can use it and what is it useful for?
Also, going back a step: “What is data?” Session participants suggested that while the public or ‘newbies’ might equate data with statistics, ‘data’ was much broader than that. It might be the raw data – or numbers – on which the stats were based. But it might also be text, or photographs.
Pretty much anything that tells somebody that something has happened; but in its raw, unanalysed form. From there, successive speakers suggested:
- “Open data is data that is available to anyone”.
- It tends to be data released by public bodies – companies can release data as open data; but they don’t often do this.
- “It has to be released under licence” and – importantly – the licence will explain what can be done with it.
The seductive allure of file formats
However, the group acknowledged, open data advocates have a tendency to jump straight into things like file formats, and to discuss open data as data that can be accessed in a particular format, such as CSV.
Nobody felt this was helpful. There was general agreement that when it came to making the case for an open data release or project it was better to talk about information and what could be done with it, or about the benefits of working in an open way.
Which doesn’t mean that when it comes to actually publishing and using open data formats and standards don’t matter. As one speaker pointed out, it would be impossible to use lightbulbs if there were no standards for how they fitted and worked; and the same applies to open data.
Aye Aye API
This lead naturally to a discussion of human and machine readable data; and to a definition of another acronym that open data experts tend to use a lot: API, or application programming interface.
“Let’s imagine Pauline is a librarian, and I want a book,” said one participant. “I go to her and fill in a card and she goes and finds the book. The API is a bit of code that’s like the librarian for your computer.”
So it’s actually all very simple? Well… there was a bit of laughter at that idea. Not least because open data releases have their own APIs; and there are some challenges to building them in the way that users might want.
— catherine brown (@CatherineB201) February 26, 2017
“As a user, you shouldn’t really be aware of the API. But there are some very complicated APIs out there, that are not very good,” one speaker said. A topic that another session at Open Data Camp 4 was due to discuss…
— Paul Armstrong (@Eighty5) February 26, 2017