Unraveling Multifaceted User Preferences on Digital Platforms: A Bayesian Deep-Learning Approach

Key Points

• The aim is to develop a Bayesian deep-learning model that effectively captures varying user preferences on digital platforms.
• Developed a Bayesian deep-learning framework
• Analyzed user activities over time
• Estimated preferences at both platform and individual levels.
• The model provides interpretable preference estimations
• Captures time-varying activities efficiently
• Demonstrates significant adaptability to user behavior changes.