Contextual explanation rules for neural clinical classifiers

Madhumita Sushil, Simon Suster, Walter Daelemans


Abstract
Several previous studies on explanation for recurrent neural networks focus on approaches that find the most important input segments for a network as its explanations. In that case, the manner in which these input segments combine with each other to form an explanatory pattern remains unknown. To overcome this, some previous work tries to find patterns (called rules) in the data that explain neural outputs. However, their explanations are often insensitive to model parameters, which limits the scalability of text explanations. To overcome these limitations, we propose a pipeline to explain RNNs by means of decision lists (also called rules) over skipgrams. For evaluation of explanations, we create a synthetic sepsis-identification dataset, as well as apply our technique on additional clinical and sentiment analysis datasets. We find that our technique persistently achieves high explanation fidelity and qualitatively interpretable rules.
Anthology ID:
2021.bionlp-1.22
Volume:
Proceedings of the 20th Workshop on Biomedical Language Processing
Month:
June
Year:
2021
Address:
Online
Venues:
BioNLP | NAACL
SIG:
SIGBIOMED
Publisher:
Association for Computational Linguistics
Note:
Pages:
202–212
Language:
URL:
https://aclanthology.org/2021.bionlp-1.22
DOI:
10.18653/v1/2021.bionlp-1.22
Bibkey:
Cite (ACL):
Madhumita Sushil, Simon Suster, and Walter Daelemans. 2021. Contextual explanation rules for neural clinical classifiers. In Proceedings of the 20th Workshop on Biomedical Language Processing, pages 202–212, Online. Association for Computational Linguistics.
Cite (Informal):
Contextual explanation rules for neural clinical classifiers (Sushil et al., BioNLP 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.bionlp-1.22.pdf
Data
MIMIC-IIISST