Mining Human Behaviors and Performance in Online Platforms

How do users of online platforms behave while fulfilling specific tasks?What are the factors that have a strong impact on a user’s performance?Is it possible to engage individuals through gamification and incentivize them to achieve better performance? In this talk, I will discuss how to model individual behaviors and groups in gamified techno-social environments, to quantify individuals’ activity and skills, and identify various cues having an impact on their performance. Our project focus on data related to some of the most popular multiplayer online games, e.g. Dota 2. Online games are indeed a modern-days natural setting in which we can both model individual behaviors and identify the mechanisms that affect users’ performance. First, I will show how to model individual trajectories via a tensor factorization model. Second, by building the underlying social network of players I will present a recommending system for teammates. Finally, I will analyze the impact that factors, such as impersonated role, mental fatigue, and social ties, have on players’ performance.

Anna Sapienza is a Postdoctoral Research associate at the USC Information Sciences Institute of Marina del Rey, CA, USA, where she works in the Machine Intelligence and Data Science group. Anna holds a Ph.D. in Applied Mathematics from the Polytechnic University of Turin, Italy, and carried out her Ph.D. studies in collaboration with the Data Science group of the ISI Foundation of Turin, Italy. Her research stays at the intersection of Data Science, Network Science, and Computational Social Science. She is particularly interested in the development of methods to monitor and model human behaviors in online techno-social environments.