A Grounded Cognitive Model for Metaphor Acquisition

Metaphors are central to our language and thought process, and modelling them computationally is imperative for reproducing human cognitive abilities. In this work, we propose a plausible grounded cognitive model for artificial metaphor acquisition. We put forward a rule-based metaphor acquisition system, which doesn’t make use of any prior ‘seed metaphor set’. Through correlation between a video and co-occurring commentaries, we show that these rules can be acquired in an unsupervised manner by an early learner capable of abstracting from multi-modal sensory input. From these grounded linguistic concepts, we derive classes based on lexico-syntactical language properties. Based on the selectional preferences of these linguistic elements, metaphorical mappings between source and target domains are acquired.

Authors

• Sushobhan Nayak Indian Institute of Technology, Kanpur
• Amitabha Mukerjee

Keywords

No keywords are indexed for this paper.