In 2016, open data was central to a growing number of projects across the globe. Throughout the year, data initiatives attempted to change the banking industry, took strides towards getting London fit, and fought “superbugs” through a real-time record of antibiotic resistance.
How will we see the open data ecosystem continue to grow in 2017? We asked the experts to tell us what the coming year will hold.
Sir Nigel Shadbolt
Principal of Jesus College, Oxford and chair and co-founder of the Open Data Institute
The UK has been top of the open data tree so far. But in 2017 we risk losing our place as others ascend, such as Australia which is opening up its map and address data and France which has recently legislated for data in its digital republic bill. I’d like to see the UK government investing in data infrastructure. Just as it builds and maintains physical infrastructure, such as transport networks, so it needs to invest in building, maintaining and opening important data, from lists of legally constituted companies, NHS procedures, environmental indicators, to maps, legal addresses, timetables and tariffs.
Too much of our data infrastructure is currently unreliable, inaccessible or only available for those who can pay. Innovators struggle to get hold of data they need, while many citizens do not feel empowered to access and use data. We must improve data skills throughout society, so policymakers, businesses and citizens can interpret and use it well.
I also want to see more algorithmic accountability: greater transparency around the algorithms that shape services and information. We should understand how software is making decisions about what we need when, in order to judge whether it is honouring our best interests. The commercial sector must be more transparent, as should government if it starts using algorithms to target and determine the public services it provides citizens.
Neil Lawrence
Professor of machine learning at University of Sheffield on leave of absence as a senior principal applied scientist at Amazon
It’s likely to be a year of consolidation for open data, we’ve seen the benefits in the scientific world, and government has bought strongly into the ideas. There remain challenges for companies in how to best extract value from open data: is it commercialisation of solutions based on open data? Or is it improvement of their internal processes in the light of open data? Or will it be something else?
A new trend is that of open models, the wide availability of pre-trained machine learning models systems which can be reused or repurposed. Some of these models could be seen as open data in distilled form. Sharing of models derived from open data might prove to be an easier way of building on the benefits. A remaining challenge is how companies exploit their own internal data sets, particularly when to obtain best value these data need to be combined with information stores from other companies.
Dr Elena Simperl
Associate professor, electronics and computer science, University of Southampton, and coordinator of ODINE (sponsors of the Guardian’s editorially independent coverage of the open data economy)
Predictions backed by lots of data and established methods have proven wrong time and time again. 2016 was the year data science failed. Engagement with data, factual evidence, and expertise is lower than ever. For open data this means one thing: that publishing it in the public domain cannot be a box-ticking exercise.
If we want these efforts to be fruitful, if we want citizens, businesses, and communities to take informed decisions using the wealth of data sources available online, we need to make this data useful and exciting for them. This can mean many things, from creating a better experience of the tools people use to explore data to publishing data in a way that people relate to.
Pavel Richter
Chief executive officer of Open Knowledge International
This year is going to be full of challenges. With restrictions on civic space taking place around the world, civil society organisations are particularly under siege. But this challenge also presents an opportunity for the open data community to foster meaningful dialogue between citizens and civil society.
2017 is the ideal time for the open data community to work with civil society and demonstrate how information can be used to drive change and build trust across divisions in society. Open data is already being used in many different contexts to improve people’s lives and we must continue to communicate how.
Openness of data in itself is not enough – we will need to engage further with users and potential users of data in 2017 to make sure the information governments and other data publishers are providing is the information citizens actually require. I look forward to seeing more collaborations between NGOs and data experts to address local, national, and transnational issues using citizen-generated data.
The Open Data Institute, the University of Southampton and the Guardian are members of the ODINE consortium.