Interpreting Motion: Grounded Representations for Spatial Language

by Inderjeet Mani, James Pustejovsky

Interpreting Motion presents an integrated perspective on how language structures constrain concepts of motion and how the world shapes the way motion is linguistically expressed. Natural language allows for efficient communication of elaborate descriptions of movement without requiring a precise specification of the motion. Interpreting Motion is the first book to analyze the semantics of motion expressions in terms of the formalisms of qualitative spatial reasoning. It shows how motion descriptions in language are mapped to trajectories of moving entities based on qualitative spatio-temporal relationships. The authors provide an extensive discussion of prior research on spatial prepositions and motion verbs, devoting chapters to the compositional semantics of motion sentences, the formal representations needed for computers to reason qualitatively about time, space, and motion, and the methodology for annotating corpora with linguistic information in order to train computer programs to reproduce the annotation. The applications they illustrate include route navigation, the mapping of travel narratives, question-answering, image and video tagging, and graphical rendering of scenes from textual descriptions. The book is written accessibly for a broad scientific audience of linguists, cognitive scientists, computer scientists, and those working in fields such as artificial intelligence and geographic information systems.

177 pages, ebook

First published February 1, 2012

About the author

Inderjeet Mani is a visiting scholar in the Department of Computer Science at Brandeis University, and senior principal scientist at the MITRE Corporation in Bedford, Massachusetts. He has also been a visiting fellow of the Computer Laboratory at the University of Cambridge, and an associate professor of linguistics at Georgetown University. He is the author or coeditor of several books, including The Language of Time and Automatic Summarization