Figurative Language Grounding in Humanoid Robots

Gwozdz, Maja (2020): Figurative Language Grounding in Humanoid Robots. In: Intelligent Systems and Applications, Vol 2, Vol. 1038: pp. 347-362

Abstract

The paper proposes a method for representing idioms (in particular, colour-related idioms) within the framework of Fluid Construction Grammar (FCG). The main purpose of this endeavour is to facilitate (figurative) language grounding in humanoid robots in the near future. There is sufficient evidence in the cognitive robotics literature that figurative language plays (or, will play) an important role in humanoid robots. Idioms pose a real challenge to satisfactory human-robot interaction due to the plethora of linguistic and contextual subtleties. In my approach, the starting point for encoding idioms within FCG was the selection of several colour-related idioms due to the fact that they could be well-integrated in current robotics experiments. Humanoid robots (for instance, the iCub) are frequently subjected to simple colour-related experiments (such as, learning how to lift and manipulate colourful objects, etc.). Introducing much more difficult tasks, such as recognising colour-idioms as abstract, is needed to achieve further robotic development. The paper presents some techniques on the theoretical encoding of idioms, as well as practical guidelines on how to implement custom figurative constructions within FCG.

• Item Type: Journal article
• Faculties: Mathematics, Computer Science and Statistics > Computer Science
• Subjects: 000 Computer science, information and general works > 004 Data processing computer science
• ISSN: 2194-5357
• Language: English