Attended by over 80 developers and students, the event was themed around Making Northern Ireland Better. Teams were tasked to identify the biggest problems affecting day to day life in Northern Ireland and come up with solutions and create something of social value, before presenting their ideas to the judging panel.
12 hours is a short turnaround by average hackathon standards, and the teams went straight in to identifying problems and thinking about potential solutions, concentrating on how to address and fix the issues before the practical aspects of their projects. That might mean that less time is left to develop an actual product, but it’s actually reassuring to see students and developers focus on problems/solutions rather than the nitty-gritty of coding.
A number of solutions were developed around the areas of homelessness, tourism and days out, sport and leisure facilities, community watchdogs, volunteering, neighbourhood participation, and more (including dog fouling).
Some points struck me about how open data can help in solving social problems and how it can be improved in doing so:
- Hackers want to solve problems, and they need data to do it
Real-world problems require real-world solutions, and for that, accurate data that represents the world is needed. That’s obvious from attending events such as this. Many of the first thoughts of the teams was ‘where can we find data for that?‘ either to help figure out and define the problem, or using data to build the solution and help people make better decisions. Some of the proposals didn’t make use of any dataset, perhaps because a relevant dataset isn’t available, instead working on gamification solutions. There’s nothing wrong with that, of course, because they’re also focusing on their users. However, access to use accurate, relevant and useful open data could bring additional benefits to those solutions by making them more relevant to the world around them as well as much as with the end-user.
- Communication and context is always important around open data
Some data sources lack comprehensive descriptions of their contents, though they might contain basic metadata (data about the dataset, but not the actual contents). This is fine in cases where the fields, values and use is obvious, but in others a bit more precision would be of great benefit to data users trying to figure out what exactly a dataset means – is that percentage or a number? Does the year relate to one day of the year or an annual average? As well as basic data descriptions, creating a two-way channel around the use of open data would help to produce better solutions. Hackathons can act as a way to achieve this, so long as data publishers attend them and listen to the feedback coming their way. But on a day-to-day basis, the ability to communicate with dataset authors or owners can also help. Getting those systems in place and getting people signed up to using them will take a bit of time, but getting the data to be used out there in the wild in the first place will help with that.
- Publishers need think how datasets can easiest be used
A good example are the coordinates systems that many of the published OpenDataNI datasets use. The standard system used in the NI public sector is the Irish Grid reference system, one of many thousands of coordinate reference systems in use across the world. The problem with this as a default is that there are only a few tools that can actually interpret this system. If the datasets are going to be useful in web-facing applications, the coordinates have to be converted into a more popular system, such as simple decimal degrees (known as WGS84, a standard format used by GPS navigation. 54.6, -5.91, for example). For the average user (and not just novices) this conversion is not straightforward. Generally, it requires specialist GIS software and an understanding of the appropriate conversion. Instead, the data holder would be much better placed to perform this conversion before producing the data. Some datasets on OpenDataNI do have the more popular global system, but many are still being published with the standard Irish Grid references. Just compare these two versions of the same thing – the dataset containing the decimal degrees coordinates have more potential to be used in applications, the version with only Irish Grid reference formats might stump some developers who aren’t used to them.
We’d given a talk to the teams about Open Data and APIs at an earlier meetup event, in particular how to use the OpenDataNI and Detail Data portals to find and query open datasets. So it was great to be able to get to the main event and to give advice and assistance to teams as they set about approaching their ideas.
Events like Hack the Hub are great for people working at the publication end of open data to get a feel for how their data is being used and to learn what they can to improve publication and dissemination. We’re looking forward to seeing many more in Northern Ireland in the near future!