I have studied how to bridge RL with NLP techniques to develop ML methods for Controlled Sequence Generation (CSG) and apply them to real-world problems. My recent interest lies in understanding the theoretical foundation of CSG.As a post-doctoral position at UNIST, I focus on developing ML algorithms grounded on RL and NLP techniques extensible to industrial applications (e.g., diet optimization, drug repurposing, sequential recommendation, and task-oriented chatbot). In addition, I study the theoretical foundations of how RL and NLP techniques get intertwined and fused through the lens of Bayesian perspectives.