Two Approaches for Generating Size Modifiers

This paper offers a solution to a small problem within a much larger problem. We focus on modelling how people use size in reference, words like "big" and "tall", which is one piece within the much larger problem of how people refer to visible objects. Examining size in isolation allows us to begin untangling a few of the complex and interacting features that affect reference, and we isolate a set of features that may be used in a hand-coded algorithm or a machine learning approach to generate one of six basic size types. The hand-coded algorithm generates a modifier type with a high correspondence to those observed in human data, and achieves 81.3% accuracy in an entirely new domain. This trails oracle accuracy for this task by just 8%. Features used by the hand-coded algorithm are added to a larger set of features in the machine learning approach, and we do not find a statistically significant difference between the precision and recall of the two systems. The input and output of these systems are a novel characterization of the factors that affect referring expression generation, and the methods described here may serve as one building block in future work connecting vision to language.

[1]  森田 純哉,et al.  33rd Annual Meeting of the Cognitive Science Society(CogSci 2011) 報告 , 2011 .

[2]  Robert Dale,et al.  Speaker-Dependent Variation in Content Selection for Referring Expression Generation , 2010, ALTA.

[3]  Emiel Krahmer,et al.  Effects of scene variation on referential overspecification , 2011, CogSci.

[4]  Kees van Deemter,et al.  Applying Machine Learning to the Choice of Size Modifiers , 2011 .

[5]  Kees van Deemter,et al.  Natural Reference to Objects in a Visual Domain , 2010, INLG.

[6]  Ian H. Witten,et al.  The WEKA data mining software: an update , 2009, SKDD.

[7]  Emiel Krahmer,et al.  Graph-Based Generation of Referring Expressions , 2003, CL.

[8]  Robert Dale,et al.  Computational Interpretations of the Gricean Maxims in the Generation of Referring Expressions , 1995, Cogn. Sci..

[9]  Ehud Reiter,et al.  Book Reviews: Building Natural Language Generation Systems , 2000, CL.

[10]  Kees van Deemter Generating Referring Expressions that Involve Gradable Properties , 2006, CL.

[11]  Werner Deutsch,et al.  Psychologie der Objektbenennung , 1976 .

[12]  Kees van Deemter,et al.  On the Use of Size Modifiers When Referring to Visible Objects , 2011, CogSci.

[13]  Péter Szigetvári,et al.  What and When? , 2019, Inauguration and Liturgical Kingship in the Long Twelfth Century.

[14]  Julie C. Sedivy,et al.  Achieving incremental semantic interpretation through contextual representation , 1999, Cognition.

[15]  Albert Gatt,et al.  Attribute Selection for Referring Expression Generation: New Algorithms and Evaluation Methods , 2008, INLG.

[16]  John K. Pate,et al.  Proceedings of the 33rd Annual Conference of the Cognitive Science Society , 2011 .

[17]  M. Tanenhaus,et al.  Watching the eyes when talking about size: An investigation of message formulation and utterance planning , 2006 .

[18]  Josef van Genabith,et al.  Dynamically structuring, updating and interrelating representations of visual and linguistic discourse context , 2005, Artif. Intell..

[19]  L. R. Dice Measures of the Amount of Ecologic Association Between Species , 1945 .

[20]  B. Landau,et al.  “What” and “where” in spatial language and spatial cognition , 1993 .