Azimuth: Systematic Error Analysis for Text Classification

TL;DR

Azimuth is an open-source tool that streamlines error analysis in text classification, helping practitioners identify and address model weaknesses to improve AI system reliability.

Contribution

It introduces a systematic approach and a comprehensive tool for error analysis in text classification, integrating multiple ML techniques in one platform.

Findings

Facilitates discovery of model weaknesses

Integrates diverse ML techniques for error analysis

Enhances reliability of text classification models

Abstract

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.