文档介绍：Workshop on Semantic Interpretation in an Actionable Context, NAACL-HLT 2012, pages 1–6, Montr ′eal, Canada, June 3-8, 2012. c2012 Association putational Linguistics Learning to Interpret Natural Language Instructions Monica Babes?-Vroman +, James MacGlashan ?, Ruoyuan Gao +, Kevin Winner ? Richard Adjogah ?,Marie desJardins ?,Michael Littman +andSmaranda Muresan ++ ?Department puter Science and Electrical Engineering University of Maryland, Baltimore County +Computer Science Department, Rutgers University ++School munication and Information, Rutgers University Abstract This paper addresses the problem of training an arti?cial agent to followverbal instructions representing high-level tasks using a set of in- structions paired with demonstration traces of appropriate behavior. From this data, a map- ping from instructions to tasks is learned, en- abling the agent to carry out new instructions in novel environments. 1 Introduction Learning to interpret language from a situated con- text has e a topic of much interest in recent years (Branavan et al., 2009; Branavan et al., 2010; Branavan et al., 2011; Clarke et al., 2010; Chen and Mooney, 2011; Vogel and Jurafsky, 2010; Gold- wasser and Roth, 2011; Liang et al., 2011; Atrzi and Zettlemoyer, 2011; Tellex et al., 2011). Instead of using annotated training data consisting of sentences and their corresponding logical forms (Zettlemoyer and Collins, 2005; Kate and Mooney, 2006; Wong and Mooney, 2007; Zettlemoyer and Collins, 2009; Lu et al., 2008), most of these approaches leverage non-linguistic information from a situated context as their primary source of supervision. These ap- proaches have been applied to various tasks such as following navigational instructions (Vogel and Ju- rafsky, 2010; Chen and Mooney, 2011; Tellex et al., 2011), software control (Branavan et al., 2009; Branavan et al., 2010), semantic parsing (Clarke et al., 2010; Liang et al., 2011) and learning to play games based on text (Branavan et al., 2