Overview

Many scientific papers and results are based on crowdsourced data. When assessing the quality, generality, bias, and ethical concerns, very little attention is paid to the specific ways in which the data was collected. However, these issues have a huge impact on the extent to which we can trust the results of the studies. In this talk, we first describe the general problem of algorithmic and AI bias and some solutions that have been identified. We then zoom in on crowdsourcing and provide a specialized discussion of the bias, ethics, and reliability challenges, as well as their proposed solutions. We conclude the talk with a call for a joint effort on reliable crowdsourcing.

Speakers

Fabio Casati is a Principal ML Engineer at ServiceNow. Fabio focuses on designing and architecting AI solutions for enterprise customers. Previously he was a Professor at the University of Trento. Prior to that, he was a technical lead for the research program on business process intelligence in Hewlett-Packard USA. He co-authored a best-selling book on Web services and is the author of over 250 peer-reviewed papers.

Fabio Casati
ServiceNow

Marcos Baez is a Senior Research Fellow at Claude Bernard University Lyon 1 where he is carrying out research on AI-enabled services. His main research interests include web engineering, crowdsourcing, human-AI interaction, and, in general, the questions of how design and engineering can be combined to make people’s lives better. Marcos has a PhD in information and communication technologies from the University of Trento.

Marcos Baez
Claude Bernard University Lyon 1

Don't miss
Don't miss our informative workshops, tutorials, and webinars.
Mon Aug 02 2021 12:19:44 GMT+0300 (Moscow Standard Time)