DOM-Q-NET: Grounded RL on Structured Language
This work addresses the problem of building efficient web navigation agents for knowledge understanding and representation learning, representing an incremental advancement with specific performance gains.
The paper tackles the challenge of web navigation for deep reinforcement learning agents by introducing DOM-Q-NET, a novel architecture that addresses large discrete action spaces and varying action counts between states, resulting in matching or outperforming existing work on the MiniWoB environment without expert demonstrations and achieving 2x improvements in sample efficiency in multi-task settings.
Building agents to interact with the web would allow for significant improvements in knowledge understanding and representation learning. However, web navigation tasks are difficult for current deep reinforcement learning (RL) models due to the large discrete action space and the varying number of actions between the states. In this work, we introduce DOM-Q-NET, a novel architecture for RL-based web navigation to address both of these problems. It parametrizes Q functions with separate networks for different action categories: clicking a DOM element and typing a string input. Our model utilizes a graph neural network to represent the tree-structured HTML of a standard web page. We demonstrate the capabilities of our model on the MiniWoB environment where we can match or outperform existing work without the use of expert demonstrations. Furthermore, we show 2x improvements in sample efficiency when training in the multi-task setting, allowing our model to transfer learned behaviours across tasks.