Multilingual-Metaphor-Detection

This page provides a fine-tuned multilingual language model XLM-RoBERTa for metaphor detection on a token-level using the Huggingface token-classification approach. Label 1 corresponds to metaphoric usage.

Reference

The training and evaluation code is available on Github. Our paper describing training and model application is available online.

@inproceedings{wachowiak2022drum,
 title={Drum Up SUPPORT: Systematic Analysis of Image-Schematic Conceptual Metaphors},
 author={Wachowiak, Lennart and Gromann, Dagmar and Xu, Chao},
 booktitle={Proceedings of the 3rd Workshop on Figurative Language Processing (FLP)},
 pages={44--53},
 year={2022}
}

Dataset

The dataset the model is trained on is the VU Amsterdam Metaphor Corpus that was annotated on a word-level following the metaphor identification protocol. The training corpus is restricted to English, however, XLM-R shows decent zero-shot performances when tested on other languages.

Results

Following the evaluation criteria from the 2020 Second Shared Task on Metaphor detection our model achieves a F1-Score of 0.76 for the metaphor-class when training XLM-RBase and 0.77 when training XLM-RLarge..

We train for 8 epochs loading the model with the best evaluation performance at the end and using a learning rate of 2e-5. From the allocated training data, 10% are utilized for validation, while the final test set is being kept separate and only used for the final evaluation.