Neural Metaphor Detection in Context

@inproceedings{Gao2018NeuralMD,
  title={Neural Metaphor Detection in Context},
  author={Ge Gao and Eunsol Choi and Yejin Choi and Luke Zettlemoyer},
  booktitle={EMNLP},
  year={2018}
}
We present end-to-end neural models for detecting metaphorical word use in context. We show that relatively standard BiLSTM models which operate on complete sentences work well in this setting, in comparison to previous work that used more restricted forms of linguistic context. These models establish a new state-of-the-art on existing verb metaphor detection benchmarks, and show strong performance on jointly predicting the metaphoricity of all words in a running text. 
End-to-End Sequential Metaphor Identification Inspired by Linguistic Theories
TLDR
By testing on three public datasets, it is found that two DNN models inspired by two human metaphor identification procedures achieve state-of-the-art performance in end-to-end metaphor identification. Expand
Improving Neural Metaphor Detection with Visual Datasets
TLDR
This work uses a straightforward technique for sampling text from Vision-Language datasets to create a data structure the authors term a visibility word embedding, and combines these embeddings in a relatively simple BiLSTM module augmented with contextualized word representations (ELMo). Expand
Being neighbourly: Neural metaphor identification in discourse
TLDR
This paper presents the first neural metaphor processing architecture that models a broader discourse through the use of attention mechanisms, and advances the state of the art on the all POS track of the 2018 VU Amsterdam metaphor identification task. Expand
Metaphor Detection using Context and Concreteness
TLDR
The model is an ensemble, utilising contextualised and static distributional semantic representations, along with word-type concreteness ratings, which predicts word metaphoricity with a deep multi-layer perceptron. Expand
Linguistic Analysis Improves Neural Metaphor Detection
TLDR
It is shown that using syntactic features and lexical resources can automatically provide additional high-quality training data for metaphoric language, and this data can cover gaps and inconsistencies in metaphor annotation, improving state-of-the-art word-level metaphor identification. Expand
Augmenting Neural Metaphor Detection with Concreteness
TLDR
This work hypothesis that concreteness and related psycholinguistic information are not available to deep neural models and that their addition will boost the performance of these models in detecting metaphor is tested and found to be true. Expand
Learning Outside the Box: Discourse-level Features Improve Metaphor Identification
TLDR
It is argued that broader discourse features are crucial for better metaphor identification and trained simple gradient boosting classifiers on representations of an utterance and its surrounding discourse learned with a variety of document embedding methods obtain near state-of-the-art results. Expand
Sensorimotor Enhanced Neural Network for Metaphor Detection
TLDR
Results show that the current model outperforms most recent works by 0.5%11% F1, indicating the effectiveness of using modality norms for metaphor detection, and provides a new perspective to detect token-level metaphoricity by leveraging the modality mismatch between words. Expand
Character aware models with similarity learning for metaphor detection
TLDR
This paper proposes to add character based word representation via a BiLSTM model and a Transformer Encoder model similar to BERT to perform metaphor identification and utilizes a similarity network to contrast the difference between literal and contextual meaning. Expand
ALBERT-BiLSTM for Sequential Metaphor Detection
TLDR
In the metaphor detection task, a sequence labeling model based on ALBERT-LSTM-softmax is designed and achieves a 0.707 F1-score for the all POS subtask and a 1.1- score for the verb subtask on the TOEFL dataset. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 36 REFERENCES
Bigrams and BiLSTMs Two Neural Networks for Sequential Metaphor Detection
TLDR
Two alternative deep neural architectures to perform word-level metaphor detection on text are presented and compared: a bi-LSTM model and a new structure based on recursive feed-forward concatenation of the input. Expand
Di-LSTM Contrast : A Deep Neural Network for Metaphor Detection
TLDR
This paper presents a deep neural architecture for metaphor detection which exploits the contrast between the contextual and general meaning of a word and uses cost-sensitive learning by re-weighting examples, and baseline features like concreteness ratings, POS and WordNet-based features. Expand
Learning to Identify Metaphors from a Corpus of Proverbs
TLDR
A novel set of features is designed to better capture the peculiar nature of proverbs and it is demonstrated that these new features are significantly more effective on the metaphorically dense proverb data. Expand
Metaphor Detection with Cross-Lingual Model Transfer
We show that it is possible to reliably discriminate whether a syntactic construction is meant literally or metaphorically using lexical semantic features of the words that participate in theExpand
Grasping the Finer Point: A Supervised Similarity Network for Metaphor Detection
TLDR
This paper presents the first deep learning architecture designed to capture metaphorical composition, and demonstrates that it outperforms the existing approaches in the metaphor identification task. Expand
Different Texts, Same Metaphors: Unigrams and Beyond
TLDR
This paper describes the development of a supervised learning system to classify all content words in a running text as either being used metaphorically or not and shows how the recall of the system can be improved over this strong baseline. Expand
Cross-Lingual Metaphor Detection Using Common Semantic Features
TLDR
The CSF - Common Semantic Features method for metaphor detection is presented, supporting the hypothesis that a CSF-based classifier can be applied across languages. Expand
Metaphor Detection with Topic Transition, Emotion and Cognition in Context
TLDR
This work presents a new approach that distinguishes literal and non-literal use of target words by examining sentence-level topic transitions and captures the motivation of speakers to express emotions and abstract concepts metaphorically. Expand
Using Language Learner Data for Metaphor Detection
TLDR
The system combines a small assertion of trending techniques, which implement matured methods from NLP and ML, and uses word embeddings from standard corpora and from corpora representing different proficiency levels of language learners in a LSTM BiRNN architecture. Expand
Neural Metaphor Detecting with CNN-LSTM Model
TLDR
This model combines CNN and LSTM layers to utilize both local and long-range contextual information for identifying metaphorical information in plain texts to extract metaphors from plain texts at word level. Expand
...
1
2
3
4
...