Statistical Approaches to Metaphor

Recent years have witnessed a growing interest in statistical and machine learning approaches to the computational treatment of metaphor. As the field of computational semantics—in particular robust parsing and lexical acquisition techniques—has progressed to the point where it is possible to accurately acquire lexical, domain, and relational information from corpora, statistical approaches have opened new avenues for robust, large-scale metaphor processing models. Approaches based on distributional association measures, vector space models, supervised learning, clustering, and LDA topic modeling1 have been proposed, bringing metaphor into the arena of statistical NLP. The parameters of these models ...

Get Metaphor now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.