What’s the big deal about big data?

Unprecedented amounts of digital information, plus advanced computing platforms to store and analyse it, bring opportunities to create value from big data – as well as certain dangers Keystone

It’s a ubiquitous buzzword, but “big data” is about to bring a wealth of opportunities to Switzerland thanks to investments in research. Efforts are also underway to control the risks.

A common definition of big data is based on “the four Vs”: volume, velocity, variety and value. In short, big data deals with extremely large amounts of digital information that are processed very rapidly, using many different sources and formats.

This definition, used by Switzerland’s Federal Data Protection and Information CommissionerExternal link (FDPIC), helps us to understand why a project trying to figure out how to update 3D models of cities in real time – a key requirement for self-driving cars – is all about big data.

An effort to develop algorithms that can compress huge amounts of digital information, while still allowing it to be analysed accurately, is a little more obvious.

Both of the projects above are among the 36 research ventures announced as part of a CHF25 million ($24.9 million) Swiss National Science Foundation National Research Programme known as NRP 75: Big DataExternal link, which was launched earlier this year.  NRP 75 spans 15 universities and will be conducted over the period 2017-2021, with first results expected in 2019.

One group of NRP 75 projects is dedicated to basic research on the information technology behind big data and addresses challenges in the “four V” dimensions. Project goals include developing better algorithms and software for faster, more efficient and more secure data processing at very large scales and across networks that can span nations, continents and the globe.

Knowledge is power

Because we live in an information society, the ability to gather and store large amounts of digital information about people – whether it concerns our physical locations, our risks of cancer, or what kinds of posts we “like” on Facebook – is a double-edged sword.

Understanding and regulating big data’s impact on society is the focus of another group of NRP 75 projects. Research topics range from how to regulate big data in trade agreements to legal and ethical questions concerning data privacy for company employees, owners of self-driving cars and medical patients.

For example, there are concerns that health insurance companies could potentially deny coverage to patients based on health-risk calculations that have been fine-tuned to an unprecedented level. Medical research data also requires special provisions for patient privacy, as well as transparency as to what the data will be used for in the future.

According to the FDPIC, in the case of big data, even personal information that has been anonymised to maintain privacy can become de-anonymised once large sets of data from different sources are merged.

“Though data may be regarded as ‘anonymous’ today, thanks to rapid technological progress and additional data sources, tomorrow it may be possible to attribute it to a specific person without much difficulty, thus potentially causing a serious breach of privacy,” the FDPIC explainsExternal link on its website.

Consumer protection

Joy Demeulemeester leads the health policy team at the FRCExternal link, French-speaking Switzerland’s consumer protection bureau. She says the FRC is already working on cases concerning data privacy with health insurance companies and consent for patients who give biological samples to biobanksExternal link for medical research.

“The biggest challenge for consumers is to remain free to give or withhold access to their data. Another challenge is developing consent forms that are very clear and understandable, without any hidden agreements,” Demeulemeester tells swissinfo.ch.

“Once you give your data, it’s impossible to get it back. So we tell our clients to read all the conditions and rules of their insurance and to be careful, because health data is very sensitive. We also work on Swiss data protection legislation and give input on what we think would be good for the people.”

Another societal concern that comes with big data is employment. With such rapid and profound change in the digital landscape, traditional jobs are changing with it, and many are expected to be taken over by artificial intelligence in the coming decades – blue and white collar alike.

Creating value – and limits

According to Christian Jensen, president of the NRP 75 steering committee, any solutions Switzerland manages to develop to these technological, legal, ethical and regulatory challenges must necessarily consider the global nature of big data itself.

“It’s fine if my data is protected in Switzerland, but what if the data traces I leave on the internet go to the US, and they create value from that data? Switzerland needs to understand that globalisation and internationalisation limit the possible actions that it can take,” Jensen tells swissinfo.ch.

But he says people should not feel like big data is an unstoppable, monolithic force. Rather, it is a tool with a lot of potential that the public can – and should – help to shape.

“I hope there can continue to be a public debate about big data and digitalisation. People need to see examples of how big data is used in different domains, and then they can say whether they find it acceptable or not,” Jensen says.

“Switzerland has a very well-educated population compared with most countries. There are very strong educational and research institutions. I think Swiss society, businesses and industries are really well prepared to benefit from big data and create value from it.”

Swiss Data Science Center

A Swiss Data Science CenterExternal link (SDSC) was also launched this year as a joint venture between the Swiss Federal Institutes of Technology in Lausanne and Zurich (EPFL and ETH Zurich), aimed at expanding data science research, education and infrastructure in Switzerland. 

With one facility in each city, the new centre will house some 40 researchers and provide an open software platform for data analytics services. A new master’s course in data science will also be offered to EPFL and ETH Zurich students through the SDSC starting this autumn.

Data dilemma

In an effort to streamline digitisation, Switzerland has been struggling to implement its digital identification system, SwissID, since 2010. 

Following in the footsteps of other European countries like Estonia and Sweden, the umbrella data management system would allow consumers to register and identify themselves for a range of services, from online shopping to tax returns and e-voting. But the system has so far failed to catch on.

A new incarnation of the initiative was launched in May this year, raising hopes for company SwissSign as well as concerns for some who say personal identification should be left to the government.