When and Why does a Model Fail? A Human-in-the-loop Error Detection Framework for Sentiment Analysis
This addresses the challenge for model developers in assessing and tracing errors in sentiment analysis models, though it is incremental as it builds on existing explainable AI methods.
The paper tackles the problem of detecting erroneous predictions in sentiment analysis models before deployment by proposing a human-in-the-loop error detection framework based on explainable features, achieving high precision in identifying errors on unseen data with limited human intervention.
Although deep neural networks have been widely employed and proven effective in sentiment analysis tasks, it remains challenging for model developers to assess their models for erroneous predictions that might exist prior to deployment. Once deployed, emergent errors can be hard to identify in prediction run-time and impossible to trace back to their sources. To address such gaps, in this paper we propose an error detection framework for sentiment analysis based on explainable features. We perform global-level feature validation with human-in-the-loop assessment, followed by an integration of global and local-level feature contribution analysis. Experimental results show that, given limited human-in-the-loop intervention, our method is able to identify erroneous model predictions on unseen data with high precision.