In NLP, models are usually evaluated by reporting single-number performance scores on a number of readily available benchmarks, without much deeper analysis. Here, we argue that - especially given the well-known fact that benchmarks often contain biases, artefacts, and spurious correlations - deeper results analysis should become the de-facto standard when presenting new models or benchmarks. We present a tool that researchers can use to study properties of the dataset and the influence of those properties on their models' behaviour. Our Text Characterization Toolkit includes both an easy-to-use annotation tool, as well as off-the-shelf scripts that can be used for specific analyses. We also present use-cases from three different domains: we use the tool to predict what are difficult examples for given well-known trained models and identify (potentially harmful) biases and heuristics that are present in a dataset.
@article{arxiv.2210.01734,
title = {Text Characterization Toolkit},
author = {Daniel Simig and Tianlu Wang and Verna Dankers and Peter Henderson and Khuyagbaatar Batsuren and Dieuwke Hupkes and Mona Diab},
journal= {arXiv preprint arXiv:2210.01734},
year = {2022}
}