The Evidence Navigator for research on COVID-19 is discussed by Professor Michael Barrett and Cambridge Judge alumnus Professor Torsten Oliver Salge (PhD 2006).
Knowledge sharing is an essential element of Cambridge Judge Business School (CJBS), and scientific knowledge sharing about COVID-19 lies behind the Evidence Navigator, a tool that a Cambridge Judge alumnus, Torsten Oliver Salge (PhD 2006) helped develop with colleagues at RWTH Aachen University in Germany.
A Professor and Co-Director of the RWTH Institute for Technology and Innovation Management, Oliver is a Fellow of Cambridge Digital Innovation (CDI), whose Academic Director is Michael Barrett, Professor of Information Systems & Innovation Studies at Cambridge Judge.
Michael recently asked Torsten to discuss the Evidence Navigator and its goals:
Why did you and your colleagues at RWTH come up with the idea of mapping the scientific evidence on COVID-19?
The COVID-19 crisis has fuelled unprecented efforts in science, industry and the broader society to build new knowledge on how to contain the virus and its health, social and economic impact. As for medical research alone, more than 200 new journal articles on COVID-19 are published each day. This is very encouraging. However, the sheer volume is likely to overwhelm the absorptive capacity of most. This is a major obstacle to the effective translation of scientific insights into evidence-based decision-making at all levels from individual researchers to national health policy.
In our opinion, what matters most for building a global learning community for COVID-19 is not only the creation and unconstrained sharing of knowledge on COVID-19, but also its curation at scale. So we sought to showcase that machine learning techniques can offer valuable support in knowledge curation, as they are able to process large volumes of text in near real-time. As such, they can be a meaningful complement to the well-established manual curation of evidence.
Who do you expect will benefit most from the use of the Evidence Navigator in the current crisis?
Our Evidence Navigator has not been designed primarily for frontline clinical staff that do their very best to help as many patients as possible to overcome COVID-19 often under highly challenging conditions. It is rather meant to support those involved in research activities on COVID-19 and those interested in connecting more closely with this evidence base – be it from the media, management or policy.
Most importantly, however, we wanted to demonstrate that machine learning and articificial intelligence have an important role to play in curating knowledge on COVID-19 – but also beyond. Despite many efforts fuelled by the COVID-19 crisis, this potential remains to be fully exploited. Our hope is hence that the current crisis will help to turn machine learning and AI into partners for knowledge curation and translation.
Can you share briefly the earlier research that led to the development of the technology underpinning the Evidence Navigator?
The Evidence Navigator was informed by two long-standing streams of research in our Institute and emerged at the intersection of both. The first goes back to my own PhD at CJBS and centers on innovation and knowledge sharing in healthcare. The second was initiated by David Antons, Professor in our Institute, and deals with text mining and bibliometric techniques as means of knowledge curation in science. We applied these techniques to map the knowledge landscape of entire journals, and we are now in the process of synthesising these insights.
As a CDI Fellow, can you reflect on the journey of this translation of research to practice which has led to a widening of your contributions to both research, practice and even policy?
I am very thankful to be part of the CDI community with its passion for leveraging digital technologies to make an impact in healthcare and beyond. The work on the Evidence Navigator and leading up to it was related to translation in three important ways. First, it constituted an early attempt to make our methods useful for others in the context of our COVID-19 crises. Second, it allowed us to make new connections ourselves – now working together with medical researchers on computational literature reviews in fields other than COVID-19. Third, it is one additional use case that demonstrates that machine learning and artificial intelligence more generally can be possible game changers for knowledge translation.
How does it work? Is it freely available? Can you share the link for people to start using it?
The Evidence Navigator combines bibliometric and text mining techniques to enable users to structure, synthesise and navigate the rapidly growing scientific evidence base on COVID-19. We invite everybody to use the Evidence Navigator and extract their own insights.