Exploring societal impact: “I see it as my mission to train the next generation of collaborative scientists.”

The Source
By: Guest contributor, Sun Nov 22 2020

New Content Item (3)


Author: Guest contributor

As part of Springer Nature’s strategic partnership with The Association of Universities in the Netherlands (VSNU), we’re interviewing researchers from a whole range of disciplines about their experiences of creating societal impact through research.

In this interview, we speak to Michel Dumontier, Distinguished Professor of Data Science and founder and Director of the Institute of Data Science at Maastricht University. He explains his views on research impact and why interdisciplinary working is one of the core principles of his institute.

Can you tell us a bit about your own research work and how it links to societal impact?

I work in the area of computational discovery science, where we think about how to represent and mine data and knowledge, with targeted applications in computational drug discovery and personalized medicine. The health research component is societally relevant and people can readily identify with the questions we tackle: how do you treat disease with a drug and how do we improve the way that people get treated and improve their outcomes? 

And linked to that societal relevance, we also look at how we can incorporate this knowledge into what we call the learning healthcare system. How do practitioners like healthcare workers tap into and contribute to this global knowledge base? How do research, clinical trial findings and practice knowledge quickly disseminate into hospitals around the world, and how do we check that that is actually helping people? 

The challenges that we have faced in finding and reusing data is what led to the development of the FAIR (Findable, Accessible, Interoperable, and Reusable) Guiding Principles – which have become the de facto guidelines across the world to promote the discovery and reuse of research data.

How easy is it to translate this type of research into healthcare practice?

What I learned when I was at Stanford is that putting your innovation into the clinic is an extremely difficult problem. If you're solely an AI researcher, you're mostly working on methods. You're likely to be looking at how to solve a health problem, but not necessarily how to make it work in the clinic. 

When working with patient data, there are all kinds of issues. There are privacy issues. There are ethical issues. There are IP issues. Essentially, there are all kinds of real ethical legal problems when working in the healthcare space. And, actually, I would argue that our healthcare system could be 100,000,000 times better, but research isn’t being translated into practical use.

What I've been doing here at Maastricht University, is looking at something we call ‘responsible data science by design’. I'm trying to devise a framework that anybody could take and immediately understand how to be successful in translating research projects into the real world. It will cover the kinds of partnerships you need, the kinds of legal support that you need, the kinds of agreements that you need for data sharing and data processing, and so on.

And this naturally leads to the idea of interdisciplinary working. There are so many open questions about applying AI technologies that it begs the question, “How do we create AI systems that, from the start, are intended to be used in society?”

To do that we need to bring together the right stakeholders – the legal scholars, the ethicists, the IT people, the data scientists, and so on, to really understand the problem we want to solve and make sure that it's feasible from day one rather than day 1,000.

You founded and are now director of the Institute of Data Science. Can you tell us about how it operates and what the benefits and challenges of its interdisciplinary approach have been?

The institute was interdisciplinary in its conception – the whole idea is that it’s a welcome place for researchers from any discipline who want to work on the problem of data in society.

From the outset we haven’t had any specific expectations about an academic’s background – whether it was computational, legal, social sciences or otherwise. If they care about these problems, then they’re a welcome addition to the institute. 

Our ambition is to bring the right people together to solve societally interesting problems, from climate change to agrifood. Instead of working within disciplinary boundaries, we ask, "What is the problem we are trying to solve and how should we best solve that problem? Who do we need to tackle this problem?" 

I see it as my mission to train the next generation of collaborative scientists. To bring all these different researchers together to speak the same speak, to talk to each other. And that can be a real challenge in and of itself. 

To give you an example, I remember in one of our early group meetings we had a technically-oriented researcher make a presentation about a system they had developed. And all the other techies were immediately onboard and excited about what they’d done. Then our legal scholar raised her hand and asked the question, "I'm sorry, but for whom was this presentation and, really, what was it about?" 

That was a classic example of how there's no shared point of understanding. But when you start having those conversations, you start to co-create a future together. That's the exciting part.

How do you go about measuring impact at the institute?

I think in any discipline or any area of research, what we should value is contributions to the field. The fundamental problem at the moment is that the metrics used to measure this aren’t really telling the whole story. 

Publishing research in conferences and journals is an important way to disseminate your work. While the number of citations for published work offers insight into its prominence, it fails to describe why the work matters to others. 

I ask my researchers not to focus on the number of citations or their H-index, but rather to articulate how others build on their research – whether that’s using the software or data they have created, or in developing new methods and their application. 

Researchers need to better characterise what I call the impact story – the consequences of having published their work. And I encourage my researchers to tell me this story when we meet in our progress meetings, annual reviews, and in making their case for tenure and promotion.

What advice would you give to early career researchers who want to have an impact and work across disciplines?

An early step for a researcher who is looking for collaborators is to look far outside of the group of people they typically spend time with. I tell our early career researchers to go to meetings that they don't present at, that they don't submit their research at, and to get onto those programme committees, talk to those organisers, and get to know those communities. 

Doing that would be an indicator for me that they’re having an impact. I want to know: what programme committees they’re on that have nothing to do with their field; what talks outside their field they’ve been invited to give, whether they're department talks, workshop talks or conference talks; and have they contacted any news organisations to tell them about their research and why it’s interesting.

Early career researchers need to figure out how to make what they’ve done as impactful as possible by communicating those results to as many people as they can. It will raise their profiles, help them find new people to work with – it’s a worthwhile investment that will pay off in the long term every time.

Communicating your research to non-specialist audiences – particularly the media and public – requires quite a specific skill set. Do you think early career researchers are given enough support to do this?

Institutions do like to promote the same people over and over again. They have their star researchers and those star researchers always get the marketing and communication attention. But there's so much talent in the university, starting even with undergraduate students.

I think universities and other organisations need to pull out their younger talent and put them in the spotlight. That’s a vital part of helping them to learn what it means to do these things at an early age and to do it well. 

Nobody gave me all the training to do the job I’m doing today, and I'm trying to change that at the institute by giving researchers more of these experiences. For instance, when I'm invited to a keynote, I try to position one of my other researchers into those positions. Or, when we’re invited to give a grant or to participate in a grant, I put them into the position of contributing to the work package or helping to organise the grant.

I'm trying to put them in a position of leadership, but then, to give them a discussion, to work with them, to mentor them, to help them understand, "What does it take to be successful in that job?" And then share that knowledge that I've accrued haphazardly so that they can do it better from the start.

Do you think all researchers need to be focused on societal impact?

I believe that, as a researcher, contributing to basic knowledge is societal impact. It's high societal impact, just as much as, for instance, collaborating with a patient advocacy group on a particular disease. Those are both contributions to society. 

People need to figure out what contributions they want to make and then how do they do that really, really well – so that every single ounce of their energy goes into delivering the results with the outcomes they want to see. 

I don't want to prejudge what that means. I just think that people need to think more clearly about what it means for them, and then to develop a plan that helps them achieve that societal impact, whatever that may be.

Learn more about how Springer Nature and VSNU are supporting researchers working to increase their societal impact 

New Content Item (3)
About Michel Dumontier

Dr. Michel Dumontier is the Distinguished Professor of Data Science at Maastricht University and co-founder of the FAIR (Findable, Accessible, Interoperable and Reusable) data principles. His research focuses on the development of computational methods for scalable and responsible discovery science.Dr. Dumontier obtained his BSc (Biochemistry) in 1998 from the University of Manitoba, and his PhD (Bioinformatics) in 2005 from the University of Toronto.

Previously a faculty member at Carleton University in Ottawa and Stanford University in Palo Alto, Dr. Dumontier founded and directs the interfaculty Institute of Data Science at Maastricht University to develop sociotechnological systems for responsible data science by design. His work is supported through the Dutch National Research Agenda, the Netherlands Organisation for Scientific Research, Horizon 2020, the European Open Science Cloud, the US National Institutes of Health and a Marie-Curie Innovative Training Network. He is the editor-in-chief for the journal Data Science and is internationally recognized for his contributions in bioinformatics, biomedical informatics, and semantic technologies including ontologies and linked data.

Follow Dr Dumontier on twitter


Author: Guest contributor

Guest Contributors include Springer Nature staff and authors, industry experts, society partners, and many others. If you are interested in being a Guest Contributor, please contact us via email: thesource@springernature.com.

Related Tags: