GrammarSHAP: An Efficient Model-Agnostic and Structure-Aware NLP Explainer
Edoardo Mosca, Defne Demirtürk, Luca Mülln, Fabio Raffagnato, Georg Groh
Abstract
Interpreting NLP models is fundamental for their development as it can shed light on hidden properties and unexpected behaviors. However, while transformer architectures exploit contextual information to enhance their predictive capabilities, most of the available methods to explain such predictions only provide importance scores at the word level. This work addresses the lack of feature attribution approaches that also take into account the sentence structure. We extend the SHAP framework by proposing GrammarSHAP—a model-agnostic explainer leveraging the sentence’s constituency parsing to generate hierarchical importance scores.- Anthology ID:
- 2022.lnls-1.2
- Volume:
- Proceedings of the First Workshop on Learning with Natural Language Supervision
- Month:
- May
- Year:
- 2022
- Address:
- Dublin, Ireland
- Editors:
- Jacob Andreas, Karthik Narasimhan, Aida Nematzadeh
- Venue:
- LNLS
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 10–16
- Language:
- URL:
- https://aclanthology.org/2022.lnls-1.2
- DOI:
- 10.18653/v1/2022.lnls-1.2
- Cite (ACL):
- Edoardo Mosca, Defne Demirtürk, Luca Mülln, Fabio Raffagnato, and Georg Groh. 2022. GrammarSHAP: An Efficient Model-Agnostic and Structure-Aware NLP Explainer. In Proceedings of the First Workshop on Learning with Natural Language Supervision, pages 10–16, Dublin, Ireland. Association for Computational Linguistics.
- Cite (Informal):
- GrammarSHAP: An Efficient Model-Agnostic and Structure-Aware NLP Explainer (Mosca et al., LNLS 2022)
- PDF:
- https://preview.aclanthology.org/ingest-acl-2023-videos/2022.lnls-1.2.pdf
- Data
- IMDb Movie Reviews, SST, SST-2