Learning Action Conditions from Instructional Manuals for Instruction Understanding
This addresses the need for autonomous agents and assistive AI to comprehend complex instructions, but it is incremental as it builds on existing NLP models with new heuristics and datasets.
The paper tackles the problem of inferring pre- and postconditions of actions from instructional manuals, which is crucial for instruction understanding in AI applications, and shows improvements of over 20% F1-score with context and over 6% with heuristics.
The ability to infer pre- and postconditions of an action is vital for comprehending complex instructions, and is essential for applications such as autonomous instruction-guided agents and assistive AI that supports humans to perform physical tasks. In this work, we propose a task dubbed action condition inference, and collecting a high-quality, human annotated dataset of preconditions and postconditions of actions in instructional manuals. We propose a weakly supervised approach to automatically construct large-scale training instances from online instructional manuals, and curate a densely human-annotated and validated dataset to study how well the current NLP models can infer action-condition dependencies in the instruction texts. We design two types of models differ by whether contextualized and global information is leveraged, as well as various combinations of heuristics to construct the weak supervisions. Our experimental results show a >20% F1-score improvement with considering the entire instruction contexts and a >6% F1-score benefit with the proposed heuristics.