Montreal NLP Meetup - "Azimuth: Systematic Error Analysis for Text Classification"

This is a recording of the presentation of "Azimuth: Systematic Error Analysis for Text Classification" at the Montreal NLP Meetup on Friday, October 21, 2022 by Orlando E. Marquez and Gabrielle Gauthier-Melançon 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 is facilitated by Azimuth. Our aim is to help AI practitioners discover and correct areas where the model does not generalize by leveraging and integrating many 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 Learn more about the project, the authors, and get the open-source code, here: https://www.servicenow.com/research/publication/gabrielle-gauthier-melancon-azim-emnlp2022.html