On the Interaction of Belief Bias and Explanations

Ana Valeria González, Anna Rogers, Anders Søgaard

    Research output: Conference Article in Proceeding or Book/Report chapterArticle in proceedingsResearchpeer-review

    Abstract

    A myriad of explainability methods have been proposed in recent years, but there is little consensus on how to evaluate them. While automatic metrics allow for quick benchmarking, it isn't clear how such metrics reflect human interaction with explanations. Human evaluation is of paramount importance, but previous protocols fail to account for belief biases affecting human performance, which may lead to misleading conclusions. We provide an overview of belief bias, its role in human evaluation, and ideas for NLP practitioners on how to account for it. For two experimental paradigms, we present a case study of gradient-based explainability introducing simple ways to account for humans' prior beliefs: models of varying quality and adversarial examples. We show that conclusions about the highest performing methods change when introducing such controls, pointing to the importance of accounting for belief bias in evaluation.
    Original languageEnglish
    Title of host publicationFindings of the Association for Computational Linguistics: ACL-IJCNLP 2021
    Number of pages13
    Place of PublicationOnline
    PublisherAssociation for Computational Linguistics
    Publication date1 Aug 2021
    Pages2930-2942
    Publication statusPublished - 1 Aug 2021

    Keywords

    • Explainability Methods
    • Evaluation Metrics
    • Belief Bias
    • Human Evaluation
    • Gradient-Based Explainability

    Fingerprint

    Dive into the research topics of 'On the Interaction of Belief Bias and Explanations'. Together they form a unique fingerprint.

    Cite this