Azimuth: Systematic Error Analysis for Text Classification
This tool addresses the problem of unreliable AI systems for AI practitioners by providing a more mature and integrated approach to error analysis, though it is incremental as it builds on existing ML techniques.
The authors tackled the lack of systematic tooling for error analysis in text classification by developing Azimuth, an open-source tool that integrates techniques like saliency maps and uncertainty analysis to help practitioners identify and address model generalization issues.
We present Azimuth, an open-source and easy-to-use tool to perform error analysis for text classification. Compared to other stages of the ML development cycle, such as model training and hyper-parameter tuning, the process and tooling for the error analysis stage are less mature. However, this stage is critical for the development of reliable and trustworthy AI systems. To make error analysis more systematic, we propose an approach comprising dataset analysis and model quality assessment, which Azimuth facilitates. We aim to help AI practitioners discover and address areas where the model does not generalize by leveraging and integrating a range of ML techniques, such as saliency maps, similarity, uncertainty, and behavioral analyses, all in one tool. Our code and documentation are available at github.com/servicenow/azimuth.