Evaluating the State-of-the-Art of End-to-End Natural Language Generation: The E2E NLG Challenge

Abstract

This paper provides a comprehensive analysis of the first shared task on End-to-End Natural Language Generation (NLG) and identifies avenues for future research based on the results. This shared task aimed to assess whether recent end-to-end NLG systems can generate more complex output by learning from datasets containing higher lexical richness, syntactic complexity and diverse discourse phenomena. Introducing novel automatic and human metrics, we compare 62 systems submitted by 17 institutions, covering a wide range of approaches, including machine learning architectures -- with the majority implementing sequence-to-sequence models (seq2seq) -- as well as systems based on grammatical rules and templates. Seq2seq-based systems have demonstrated a great potential for NLG in the challenge. We find that seq2seq systems generally score high in terms of word-overlap metrics and human evaluations of naturalness -- with the winning SLUG system (Juraska et al., 2018) being seq2seq-based. However, vanilla seq2seq models often fail to correctly express a given meaning representation if they lack a strong semantic control mechanism applied during decoding. Moreover, seq2seq models can be outperformed by hand-engineered systems in terms of overall quality, as well as complexity, length and diversity of outputs. This research has influenced, inspired and motivated a number of recent studies outwith the original competition, which we also summarise as part of this paper.

Computer Speech and Language, final accepted manuscript (in press)

References in corpus (9)

• Sequence to Sequence Learning with Neural Networks
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• Microsoft COCO Captions: Data Collection and Evaluation Server
• Learning to Search Better Than Your Teacher
• Attention-Based Recurrent Neural Network Models for Joint Intent Detection and Slot Filling
• Challenges in Data-to-Document Generation
• Referenceless Quality Estimation for Natural Language Generation
• Crowd-sourcing NLG Data: Pictures Elicit Better Data
• Constrained Decoding for Neural NLG from Compositional Representations in Task-Oriented Dialogue

Cited by in corpus (15)

• The GEM Benchmark: Natural Language Generation, its Evaluation and Metrics
• Anticipating Safety Issues in E2E Conversational AI: Framework and Tooling
• Metaphoric Paraphrase Generation
• Learning to Reason for Text Generation from Scientific Tables
• Neural Data-to-Text Generation via Jointly Learning the Segmentation and Correspondence
• Logical Natural Language Generation from Open-Domain Tables
• Improved Natural Language Generation via Loss Truncation
• Constrained Decoding for Neural NLG from Compositional Representations in Task-Oriented Dialogue
• RevCore: Review-augmented Conversational Recommendation
• Towards Faithfulness in Open Domain Table-to-text Generation from an Entity-centric View
• Document Automation Architectures and Technologies: A Survey
• Partially-Aligned Data-to-Text Generation with Distant Supervision
• DORB: Dynamically Optimizing Multiple Rewards with Bandits
• AGGGEN: Ordering and Aggregating while Generating
• Let Your Heart Speak in its Mother Tongue: Multilingual Captioning of Cardiac Signals