Human Intent
Understanding human intent is a crucial area of research aiming to enable machines to better interact with and assist humans. Current efforts focus on inferring intent from various modalities, including language, visual cues (gaze, body posture), and sensor data (accelerometers), often employing deep learning models like transformers and graph neural networks, along with techniques such as plan recognition and Markov Decision Processes. This research is significant for advancing human-computer interaction, improving the safety and efficiency of autonomous systems (e.g., robots, self-driving cars), and enhancing applications like recommender systems and content moderation by aligning with user needs and ethical considerations.
Papers
5IDER: Unified Query Rewriting for Steering, Intent Carryover, Disfluencies, Entity Carryover and Repair
Jiarui Lu, Bo-Hsiang Tseng, Joel Ruben Antony Moniz, Site Li, Xueyun Zhu, Hong Yu, Murat Akbacak
Hierarchical Reinforcement Learning for Modeling User Novelty-Seeking Intent in Recommender Systems
Pan Li, Yuyan Wang, Ed H. Chi, Minmin Chen