Idiomatic Expression Identification using Semantic Compatibility

Abstract

Idiomatic expressions are an integral part of natural language and constantly being added to a language. Owing to their non-compositionality and their ability to take on a figurative or literal meaning depending on the sentential context, they have been a classical challenge for NLP systems. To address this challenge, we study the task of detecting whether a sentence has an idiomatic expression and localizing it when it occurs in a figurative sense. Prior research for this task has studied specific classes of idiomatic expressions offering limited views of their generalizability to new idioms. We propose a multi-stage neural architecture with attention flow as a solution. The network effectively fuses contextual and lexical information at different levels using word and sub-word representations. Empirical evaluations on three of the largest benchmark datasets with idiomatic expressions of varied syntactic patterns and degrees of non-compositionality show that our proposed model achieves new state-of-the-art results. A salient feature of the model is its ability to identify idioms unseen during training with gains from 1.4% to 30.8% over competitive baselines on the largest dataset.

SemEval-2022 Task 2: Multilingual Idiomaticity Detection and Sentence Embedding

“…Apart from these competitions, various studies have addressed different tasks on MWEs and their compositionality, such as: classifying verb-particle constructions (Cook and Stevenson, 2006), identifying light verb constructions and determining the literality of noun compounds (Shwartz and Dagan, 2019), identifying and classifying idioms in running text (Zeng and Bhat, 2021), as well as predicting the compositionality of several types of MWEs (Lin, 1999;McCarthy et al, 2003;Reddy et al, 2011;Schulte im Walde et al, 2013;Salehi et al, 2015).…”

“…On the one hand, some studies suggest that pre-training based on masked language modeling does not properly encode idiomaticity in word representations (Nandakumar et al, 2019;Garcia et al, 2021b,a). However, as these embeddings encode contextual information, supervised approaches using these representations tend to obtain better results in different tasks dealing with (non-)compositional semantics (Shwartz and Dagan, 2019;Fakharian and Cook, 2021;Zeng and Bhat, 2021).…”

HIT at SemEval-2022 Task 2: Pre-trained Language Model for Idioms Detection

“…So far, there has been extensive research about idioms detection. Zeng and Bhat (2021) propose a multi-stage neural architecture with attention flow. Garcia et al (2021a,b) probs idiomaticity in vector space and propose NCTTI dataset.…”

HYU at SemEval-2022 Task 2: Effective Idiomaticity Detection with Consideration at Different Levels of Contextualization

“…raises the question of how properly to deal with idiomatic aspects of linguistic expressions (Garcia et al, 2021a,b;Zeng and Bhat, 2021). An intuitive solution for mitigating the aforementioned problem is an introduction of a sophisticated method designed to estimate the idiomaticity of a given expression, which enables the separate processing of the expression according to its category.…”