Curious Case of Language Generation Evaluation Metrics: A Cautionary Tale
This work highlights a cautionary tale for the NLP community about the risks of relying on flawed metrics, urging more careful evaluation practices.
The paper demonstrates that popular automatic evaluation metrics for language generation tasks, such as image captioning and machine translation, have critical flaws, including preferring system outputs over human texts and insensitivity to rare word translations, based on experiments across multiple datasets and tasks.
Automatic evaluation of language generation systems is a well-studied problem in Natural Language Processing. While novel metrics are proposed every year, a few popular metrics remain as the de facto metrics to evaluate tasks such as image captioning and machine translation, despite their known limitations. This is partly due to ease of use, and partly because researchers expect to see them and know how to interpret them. In this paper, we urge the community for more careful consideration of how they automatically evaluate their models by demonstrating important failure cases on multiple datasets, language pairs and tasks. Our experiments show that metrics (i) usually prefer system outputs to human-authored texts, (ii) can be insensitive to correct translations of rare words, (iii) can yield surprisingly high scores when given a single sentence as system output for the entire test set.