Abstract
Relation classification models are conventionally evaluated using only a single measure, e.g., micro-F1, macro-F1 or AUC. In this work, we analyze weighting schemes, such as micro and macro, for imbalanced datasets. We introduce a framework for weighting schemes, where existing schemes are extremes, and two new intermediate schemes. We show that reporting results of different weighting schemes better highlights strengths and weaknesses of a model.- Anthology ID:
- 2022.nlppower-1.4
- Volume:
- Proceedings of NLP Power! The First Workshop on Efficient Benchmarking in NLP
- Month:
- May
- Year:
- 2022
- Address:
- Dublin, Ireland
- Venue:
- nlppower
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 32–41
- Language:
- URL:
- https://aclanthology.org/2022.nlppower-1.4
- DOI:
- 10.18653/v1/2022.nlppower-1.4
- Cite (ACL):
- David Harbecke, Yuxuan Chen, Leonhard Hennig, and Christoph Alt. 2022. Why only Micro-F1? Class Weighting of Measures for Relation Classification. In Proceedings of NLP Power! The First Workshop on Efficient Benchmarking in NLP, pages 32–41, Dublin, Ireland. Association for Computational Linguistics.
- Cite (Informal):
- Why only Micro-F1? Class Weighting of Measures for Relation Classification (Harbecke et al., nlppower 2022)
- PDF:
- https://preview.aclanthology.org/ingestion-script-update/2022.nlppower-1.4.pdf
- Code
- dfki-nlp/weighting-schemes-report
- Data
- DocRED