How to make transparency and accountability of algorithmic systems more “real”. The session discusses constructively about what is needed to foster research at goverment level, how to improve the dissemination of best practices and software in society.
In today’s world, we can see how the rising benefits from Big Data and AI technologies have a wide impact on our economy and social organization. Transparency and trust of such Algorithmic Systems, is turning into Competitiveness Factors for this new Data-Driven Economy. Data analytics is changing from description of the past, to predictive and descriptive analytics for decision support. There are some challenges to be faced. For instance, it is often assumed that big data techniques are unbiased because of the scale of the data, and because the techniques are implemented through algorithmic systems.
However, there are many efforts to undermine these challenges such as TransAlgo, which is the National Scientific Platform for Transparency & Accountabily. TransAlgo is actively working on these challenges by being a resource centre with reports, publications, software etc., bringing awareness with workshops and Moocs.
Jun (Luke) Huan, program director at National Science Foundation and professor at The University of Kansas explained how we can we create transparency models based on the Human Constructivism Learning theory to improve interpretability and transparency models.
Judith Herzog and lofred Madzou, policymakers at Le Conseil National du Numérique, described how they are facing the challenges of speed of digital cycles and the diversity and opacity of online platforms while abiding to the principle of Accountability and law.