Woodley Packard

Papers

Total Citations

H-Index

About

Woodley Packard is a researcher whose work sits at the intersection of computational linguistics and robotics, focusing on how machines can understand and execute human spatial commands. His most-cited paper, "UW-MRS: Leveraging a Deep Grammar for Robotic Spatial Commands" (2014), introduced a novel approach to the SemEval-2014 Task 6, a challenge in context-informed supervised parsing. Packard’s key contribution was a hand-built, rule-based system that leveraged a pre-existing, broad-coverage deep grammar of English—the Minimal Recursion Semantics (MRS) framework. This allowed for precise, semantically rich interpretation of spatial language, enabling robots to understand nuanced instructions like "put the cup to the left of the box." While his citation count (7) reflects a specialized niche, his work is foundational for bridging the gap between deep linguistic analysis and practical robotic control. Packard’s approach demonstrated that deep, grammar-driven parsing could outperform data-hungry statistical methods in constrained, high-stakes domains. His research remains a touchstone for those exploring how formal semantics can ground natural language in robotic action, offering a principled alternative to end-to-end learning.