Augmenting Neural Metaphor Detection with Concreteness

Abstract

The idea that a shift in concreteness within a sentence indicates the presence of a metaphor has been around for a while. However, recent methods of detecting metaphor that have relied on deep neural models have ignored concreteness and related psycholinguistic information. We hypothesis that this information is not available to these models and that their addition will boost the performance of these models in detecting metaphor. We test this hypothesis on the Metaphor Detection Shared Task 2020 and find that the addition of concreteness information does in fact boost deep neural models. We also run tests on data from a previous shared task and show similar results.

Features of Perceived Metaphoricity on the Discourse Level: Abstractness and Emotionality

“…After all, research in cognitive linguistics views metaphor as a figurative device for transferring knowledge from a concrete domain to a more abstract domain (Lakoff and Johnson, 1980), as exemplified by the conceptual metaphor LIFE IS A JOURNEY in the introduction. This view led to the hypothesis that metaphorical word usage is correlated with the degree of abstractness of the word's contexts, and this hypothesis has been supported in numerous studies where abstractness features have proven useful for automated metaphor identification (Turney et al, 2011;Tsvetkov et al, 2013;Köper and Schulte im Walde, 2016b;Alnafesah et al, 2020;Hall Maudslay et al, 2020).…”

“…The vast majority of work in NLP has been concerned with the detection of metaphorical expres-sions 1 . Linguistically and conceptually-driven strategies involve identifying selectional preference violations (compare consume food vs. consume information) (Fazly et al, 2009;Shutova et al, 2013;Ehren et al, 2020), judging discourse coherence (Sporleder and Li, 2009;Bogdanova, 2010;Dankers et al, 2020), and inducing discourse features indicating figurative language, such as supersenses, concreteness, emotionalty, imageability (Turney et al, 2011;Tsvetkov et al, 2013;Köper and Schulte im Walde, 2016b;Mohammad et al, 2016;Köper and Schulte im Walde, 2018;Alnafesah et al, 2020;Hall Maudslay et al, 2020). Existing research has however focused solely on sentences with word-level and phrase-level metaphorical expressions, and there has been little discussion on metaphoricity on the discourse level.…”

Neural Metaphor Detection with Visibility Embeddings

“…Impressive results 1 were presented in the 2018 Metaphor Detection Shared Task (Leong et al, 2018), with most of the groups using neural models with other linguistic elements like POS tags, Word-Net features, concreteness scores and more (Wu et al, 2018;Swarnkar and Singh, 2018;Pramanick et al, 2018;Bizzoni and Ghanimifard, 2018), as well as in the more recent 2020 Shared Task , with the majority of groups using some variation of BERT in addition to the other features Gao and Zhang, 2002;Kuo and Carpuat, 2020;Torres Rivera et al, 2020;Kumar and Sharma, 2020;Hall Maudslay et al, 2020;Stemle and Onysko, 2020;Liu et al, 2020;Brooks and Youssef, 2020;Alnafesah et al, 2020;Wan et al, 2020;Dankers et al, 2020).…”

Modality Enriched Neural Network for Metaphor Detection

“…Although 90% of the systems adopt deep learning architectures (Dankers et al, 2019;Gao et al, 2018;Gutierrez et al, 2017), there are substantial linguist resources employed for system refinement, such as data with information of concreteness and imageability, semantic encoding of WordNet, FrameNet, VerbNet, SUMO ontology, property norms, syntactic dependency patterns, sensorial and vision-based information, etc. Such external resources provide different linguistic cues for metaphoricity inference, demonstrating their effectiveness for metaphor detection to various degrees (Alnafesah et al, 2020;Klebanov et al, 2016;.…”

“…Early studies of metaphor detection tend to adopt feature-engineering in a supervised machine learning paradigm, which construct feature vectors based on concreteness and imageability, semantic classification using WordNet, FrameNet, VerbNet, SUMO ontology, property norms and distributional semantic models, syntactic dependency patterns, sensorial and vision-based features (Alnafesah et al, 2020;Klebanov et al, 2016;Gutierrez et al, 2016).…”