My research

I am a tenured researcher at CWI, the National Research Institute for Mathematics and Computer Science in the Netherlands, in the Human-Centered Data Analytics group. My research focuses on the transparent and explainable prediction of the quality of the information online. I develop AI pipelines that, by combining machine learning, symbolic reasoning, and human computation, aim at (semi-)automating the assessment process. On the one hand, these pipelines aim at predicting the quality of information items most accurately and extensively, considering that it is important to consider multiple aspects of quality to serve the needs of multiple subjects in different contexts. On the other hand, it is important to make these pipelines transparent in order to foster trust in the resulting assessments and favor conscious intervention in the pipelines when users need to tune them.

IAS fellowship

In the context of the IAS/DSC fellowship, I will further investigate the links between the quality of information items and their logical and argumentative structure. Using crowdsourcing and human computation, I will first focus on mining arguments from text and their types, as defined by the "Periodic Table of Arguments" (Wagemans, 2016; Hinton & Wagemans, 2021). This will allow refining the evaluation of how such arguments are used, their strength, and the possible correlation with the quality of the items of information that show them.