Towards Faithfulness in Open Domain Table-to-text Generation from an Entity-centric View

Abstract

In open domain table-to-text generation, we notice that the unfaithful generation usually contains hallucinated content which can not be aligned to any input table record. We thus try to evaluate the generation faithfulness with two entity-centric metrics: table record coverage and the ratio of hallucinated entities in text, both of which are shown to have strong agreement with human judgements. Then based on these metrics, we quantitatively analyze the correlation between training data quality and generation fidelity which indicates the potential usage of entity information in faithful generation. Motivated by these findings, we propose two methods for faithful generation: 1) augmented training by incorporating the auxiliary entity information, including both an augmented plan-based model and an unsupervised model and 2) training instance selection based on faithfulness ranking. We show these approaches improve generation fidelity in both full dataset setting and few shot learning settings by both automatic and human evaluations.

AAAI-21 Camera Ready

References in corpus (9)

• Evaluating the State-of-the-Art of End-to-End Natural Language Generation: The E2E NLG Challenge
• fairseq: A Fast, Extensible Toolkit for Sequence Modeling
• Asking and Answering Questions to Evaluate the Factual Consistency of Summaries
• Strategies for Structuring Story Generation
• Shaping the Narrative Arc: An Information-Theoretic Approach to Collaborative Dialogue
• Learning to generate one-sentence biographies from Wikidata
• Order-Planning Neural Text Generation From Structured Data
• Neural Data-to-Text Generation via Jointly Learning the Segmentation and Correspondence
• Question Answering as an Automatic Evaluation Metric for News Article Summarization