Vi använder oss av cookies för att göra din upplevelse bättre.

Interview with Barend Mons

We´re very excited to have this opportunity to ask you a few questions before you arrive at MYiHealth to hold the lecture Go FAIR, Go Anywhere - The Personal Health Train on Track.

FAIR data stands for Findable, Accessible, Interoperable & Reusable. An important step in the FAIR Data approach is to publish existing and new datasets in a semantically interoperable format that can be understood by computer systems. What gave you this idea?
The idea is not new, for instance, I published in 2005 (see below) a paper that started with the statement that 'Computational Biology needs computer-readable information records'.

Have things gone as expected?
So in fact it took more than 10 years before the FAIR principles were formulated to achieve machine readability of data. Was this expected? In hindsight people tell me that this is a rather average time for a disruptive idea to gain ground and reach a tipping point.

What is your view on the future of FAIR Data?
I imagine a 'Internet of FAIR data and services' where analytical tools can find data, as well as the needed compute to enable the meta-analysis of 'big' data routinely and in a completely distributed fashion, where data essentially 'stays where it is'.

You have also chaired the project, Open Science Data Cloud. Can you tell us a little bit about it?
I chaired the High Level Expert Group for the European Open Science Cloud. This is not really a 'project', it is a vision and our report can be read on the web.

In June 2015, Carlos Moedas said that he wanted to see the results within one year. How has it gone?
The Commisioner wanted the results of the group within one year, not a fully functional EOSC, which is a goal rather for the end of the H2020 programme.

During the two years that the project has been underway, have you implemented FAIR data in this project?
In your view, what are the challenges facing Open Science Data Cloud?
The challenges are all listed in the report

PHT (Personal Health Train), which basically entails gathering everyone’s medical data, is another project that you have headed. Would you like to tell us more?
The term 'gathering' is confusing. The PHT avoids the need to 'gather' data. The compute goes to the data (see the movie). In that sense the PHT is an early implementation of the internet of FAIR data and services model.

How can you guarantee security?
The data never leaves the personal locker of people, and people decide who can enter their locker to learn on their data. Data will be also encrypted.

Do you think that you can convince people to trust it? And what will you do to earn our trust?

As we do not gather data, the trust is at the level of individual 'trains' (VM learning algorithms) and you decide whether you let trains in or not.

Barend Mons's biography »