Objective
"Metaphor is a focus in many research fields, and is important in much mundane discourse: issues such as relationships, disease, states of mind, are often most economically and understandably conveyed, asked about, discussed, etc., through metaphor. Despite its importance, there is currently a lack of work on technological solutions to modeling metaphor, with modeling difficulties stemming from its richness and complexity. We propose to meet this gap by building a prototype natural language generation and understanding system, focusing on metaphors for the domains of illness and reconciliation (e.g. illness ""attacks"" the body, ""breaking down barriers""). The resulting prototype will be capable of choosing: whether to use metaphor at all; particular metaphorical conceptions; and particular metaphorical wording. Our respective research backgrounds will enable a combination of existing approaches to modeling metaphor in AI, Cognitive Science and Formal Linguistics, which is unique within Europe. Specifically, for processing metaphor at a conceptual level we employ a current AI system, ATT-Meta, to extend Embodied Construction Grammar (ECG), a system for interfacing conceptual and linguistic levels, integrating ATT-Meta's unique capabilities for modeling metaphor into ECG. We then utilise the generation capabilities of an implemented system for modeling the linguistic level, Dynamic Syntax (DS), which uniquely models generation as fully incremental and context-dependent, using DS to model formulaic expressions of metaphorical language. By thus combining AI inference with formal linguistic and data-oriented techniques, the project will make a major contribution to metaphor and generation research within the European Research Area, yielding a system having the novel capacity of not only being able to model a range of typical metaphorical expressions, but also how these can be extended in context-dependent ways (e.g. illness ""agressively attacks"")."
Fields of science (EuroSciVoc)
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
You need to log in or register to use this function
Call for proposal
FP7-PEOPLE-2012-IIF
See other projects for this call
Funding Scheme
MC-IIF - International Incoming Fellowships (IIF)Coordinator
B15 2TT Birmingham
United Kingdom