Carlson School of Management
Twin Cities
This project focuses on developing insights on how players in online team-based games churn and what makes them more engaged, especially with the assistance of AI bots, leveraging a large, granular dataset on a team-based sports game. The factors investigated include (but are not limited to): whether an AI-bot-based teammate/opponent leads to better onboarding performance; how best to match the players in a team to improve the game experience and player loyalty; how to model player dynamics and capture player preference; and what determines the optimal levels of challenges for players at any given time. The project will not only use big data in terms of volume (as the dataset is 50 gigabytes and growing), but also leverages advanced statistical and machine-learning models (e.g. deep learning models for sequence data; NLP tasks) that require intensive computing power.