The GEM Benchmark: Natural Language Generation, its Evaluation and Metrics

Abstract

We introduce GEM, a living benchmark for natural language Generation (NLG), its Evaluation, and Metrics. Measuring progress in NLG relies on a constantly evolving ecosystem of automated metrics, datasets, and human evaluation standards. Due to this moving target, new models often still evaluate on divergent anglo-centric corpora with well-established, but flawed, metrics. This disconnect makes it challenging to identify the limitations of current models and opportunities for progress. Addressing this limitation, GEM provides an environment in which models can easily be applied to a wide set of tasks and in which evaluation strategies can be tested. Regular updates to the benchmark will help NLG research become more multilingual and evolve the challenge alongside models. This paper serves as the description of the data for which we are organizing a shared task at our ACL 2021 Workshop and to which we invite the entire NLG community to participate.

References in corpus (7)

• Multilingual Denoising Pre-training for Neural Machine Translation
• Evaluating the State-of-the-Art of End-to-End Natural Language Generation: The E2E NLG Challenge
• NUBIA: NeUral Based Interchangeability Assessor for Text Generation
• MultiWOZ 2.2 : A Dialogue Dataset with Additional Annotation Corrections and State Tracking Baselines
• QuestEval: Summarization Asks for Fact-based Evaluation
• Beyond Leaderboards: A survey of methods for revealing weaknesses in Natural Language Inference data and models
• ORB: An Open Reading Benchmark for Comprehensive Evaluation of Machine Reading Comprehension

Cited by in corpus (9)

• AMMUS : A Survey of Transformer-based Pretrained Models in Natural Language Processing
• Dynaboard: An Evaluation-As-A-Service Platform for Holistic Next-Generation Benchmarking
• Automatic Construction of Evaluation Suites for Natural Language Generation Datasets
• The Benchmark Lottery
• How to Evaluate Your Dialogue Models: A Review of Approaches
• Benchmarking Multimodal AutoML for Tabular Data with Text Fields
• Human Evaluation of Creative NLG Systems: An Interdisciplinary Survey on Recent Papers
• Datasets: A Community Library for Natural Language Processing
• Truth-Conditional Captioning of Time Series Data