|TOCCHETTI ANDREA||Cycle: XXXV |
Section: Computer Science and Engineering
Tutor: TANCA LETIZIA
Advisor: BRAMBILLA MARCO Major Research topic
:Explainability of Analysis of Human-generated Content and Behavior for Socio-political PhenomenaAbstract:
This Ph.D. focuses on understanding and explaining complex phenomena, mainly covering the social and political aspects. Furthermore, understanding which factors influence the aforementioned phenomena and how the latter evolve based on these influences are two of the main focus aspects. One of the most problematic aspects of most social phenomena is their unpredictability. We are humans after all, and most of the time it’s difficult to identify a clear pattern in our actions. In order to achieve such a result, a combination of Social Sciences and Data Science is needed. Social Data Science is a new discipline that focuses on the application of Data Science in the social field, linking big data and data analysis to the human world. Through this link, Social Data Science becomes able to understand human behavior and everyday life patterns. In the end, the result this Ph.D. aims to achieve is the development of an AI able to understand human behavior from human-generated content, automating the process of data collection, data analysis, and understanding, providing a complete and understandable outcome. Another objective this Ph.D. would like to achieve is to make the decision process of the AI as understandable as possible for humans, focusing on the analyses and the decision-making process, therefore aiming to remove the concept of “black box” from the final system. In particular, as the decision process becomes more and more understandable for humans, their feedback can become relevant during the learning process of the AI. Therefore, human intervention in the learning process could be proven to be useful, due to the aforementioned complexities the AI should deal with.