End-to-End Content and Plan Selection for Data-to-Text Generation

By Sebastian Gehrmann, Falcon Z. Dai, Henry Elder, Alexander M. Rush. 2018

73 citations recorded by the source corpus.

Learning to generate fluent natural language from structured data with neural networks has become an common approach for NLG. This problem can be challenging when the form of the structured data varies between examples. This paper presents a survey of several extensions to sequence-to-sequence models to account for the latent content selection process, particularly variants of copy attention and coverage decoding. We further propose a training method based on diverse ensembling to encourage models to learn distinct sentence templates during training. An empirical evaluation of these techniques shows an increase in the quality of generated text across five automated metrics, as well as human evaluation.

Subfields: Artificial Intelligence, Natural Language Processing, Natural Language Generation

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

In Natural Language Generation

Citations in the corpus

Neural Generation for Czech: Data and Baselines (2019, background): “Delexicalization or an analogous technique, such as a copy mechanism (Gu et al., 2016; Gehrmann et al., 2018), is required for most generation scenarios to allow generalization to unseen entity names: sets of entities are open (potentially infinite and subject to change) while training data is scarc”
DART: Open-Domain Structured Data Record to Text Generation (2020, methodology): “Recently, neural encoder-decoder models based on attention and copy mechanisms have shown promising results (Gehrmann et al., 2018; Puduppully et al., 2018, 2019; Castro Ferreira et al., 2019).”
ToTTo: A Controlled Table-To-Text Generation Dataset (2020, methodology): “While originally designed for summarization it is commonly used in data-to-text as well (Gehrmann et al., 2018).”
Revisiting Challenges in Data-to-Text Generation with Fact Grounding (2020, background): “Besides, multi-reference has been explored for evaluating data-to-text generation systems (Novikova et al., 2017) and for content selection and planning (Gehrmann et al., 2018).”
Controlling hallucinations at word level in data-to-text generation (2021, unknown): “…Juraska et al. [21] leverage templating and hand-crafted rules to re-rank the top outputs of a model decoding via beam search; Gehrmann et al. [17] also leverage the possible templating formats of E2E’s reference texts, and train an ensemble of decoders where each decoder is associated to one…”
De-Confounded Variational Encoder-Decoder for Logical Table-to-Text Generation (2021, background): “It is worth to mention that the copy mechanism (Luong et al., 2015) is an important part to deal with the out-of-vocabulary (OOV) words (Lebret et al., 2016; Gehrmann et al., 2018; Chen et al., 2020a) when not using pretrained language models.”
Aspect-Controllable Opinion Summarization (2021, background): “Text summarization models, both extractive (Narayan et al., 2018; Zheng and Lapata, 2019; Ca-chola et al., 2020) and abstractive (See et al., 2017; Gehrmann et al., 2018; Liu and Lapata, 2019), operate under the assumption that salient content is”
Evaluating Semantic Accuracy of Data-to-Text Generation with Natural Language Inference (2020, background): “State-of-the-art neural D2T models are prone to omitting or halluci-nating facts (Gehrmann et al., 2018; Castro Ferreira et al., 2019; Dušek et al., 2020), which restricts their real-world deployment.”
ESPRIT: Explaining Solutions to Physical Reasoning Tasks (2020, background): “, 2016) have shown promising results on table-to-text tasks (Wiseman et al., 2017; Gehrmann et al., 2018; Puduppully et al., 2019a,b; Iso et al., 2019; Castro Ferreira et al., 2019; Zhao et al., 2020; Chen et al., 2020a).”
Using natural language generation to bootstrap missing Wikipedia articles: A human-centric perspective (2022, unknown): “There is a large body of work that uses the encoderdecoder framework from machine translation [12,82] for NLG [11,21,22,51,55,59,79,84,90,91].”
Plan-then-Generate: Controlled Data-to-Text Generation via Planning (2021, methodology): “Gehrmann et al. (2018), Puduppully et al. (2019a,b), and Chen et al. (2020b) adopted copy mechanism for content selection to improve the information coverage of the outputs.”
Towards Faithful Neural Table-to-Text Generation with Content-Matching Constraints (2020, methodology): “Extended transformer has been applied to game summary (Gong et al., 2019) and E2E NLG tasks (Gehrmann et al., 2018).”
Semi-Supervised Neural Text Generation by Joint Learning of Natural Language Generation and Natural Language Understanding Models (2019, background): “…consen-sual NLG pipeline (Gatt and Krahmer, 2018), but there has been recently a strong interest in End-to-End (E2E) NLG systems which are able to jointly learn sentence planning and surface realization (Duˇsek and Jurc´ıcek, 2016; Agarwal et al., 2018; Juraska et al., 2018; Gehrmann et al., 2018).”
Neural data-to-text generation: A comparison between pipeline and end-to-end architectures (2019, background): “…result, in NLG more recently, neural end-to-end data-to-text models have been proposed, which directly learn input-output mappings and rely much less on explicit intermediate representations (Wen et al., 2015; Duˇsek and Jurcicek, 2016; Mei et al., 2016; Lebret et al., 2016; Gehrmann et al., 2018).”