Learning to generate one-sentence biographies from Wikidata

By Andrew Chisholm, Will Radford, Ben Hachey. 2017

112 citations recorded by the source corpus.

We investigate the generation of one-sentence Wikipedia biographies from facts derived from Wikidata slot-value pairs. We train a recurrent neural network sequence-to-sequence model with attention to select facts and generate textual summaries. Our model incorporates a novel secondary objective that helps ensure it generates sentences that contain the input facts. The model achieves a BLEU score of 41, improving significantly upon the vanilla sequence-to-sequence model and scoring roughly twice that of a simple template baseline. Human preference evaluation suggests the model is nearly as good as the Wikipedia reference. Manual analysis explores content selection, suggesting the model can trade the ability to infer knowledge against the risk of hallucinating incorrect information.

Subfields: Artificial Intelligence, Communication, Natural Language Processing

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 Communication

In Natural Language Processing

Citations in the corpus

Structure-to-Text Generation with Self-Training, Acceptability Classifiers and Context-Conditioning for the GEM Shared Task (2021, methodology): “Most approaches to self-training for NLG— including earlier work on automatic data cleaning— make use of cycle consistency between parsing and generation models (Chisholm et al., 2017; Nie et al., 2019; Kedzie and McKeown, 2019; Qader et al., 2019).”
Survey of Hallucination in Natural Language Generation (2022, background): “Data-to-Text Generation is the task of generating natural language descriptions conditioned on structured data [87, 119], such as tables [133, 191], database records [25], and knowledge graphs [54].”
Challenges in Data-to-Document Generation (2017, background): “Within the world of neural text generation, some recent work has focused on conditioning language models on tables (Yang et al., 2016), and generating short biographies from Wikipedia Tables (Lebret et al., 2016; Chisholm et al., 2017).”
Neural Wikipedian: Generating Textual Summaries from Knowledge Base Triples (2017, result): “Our task is most similar to recent work by Lebret et al. and Chisholm et al., who both employ adaptations of the encoder-decoder framework in order to generate the ﬁrst sentence of a Wikipedia biography [10, 23].”
Using natural language generation to bootstrap missing Wikipedia articles: A human-centric perspective (2022, methodology): “All these approaches use structured data from Freebase, Wikidata and DBpedia as input and generate summaries consisting either by one or two sentences that match the style of the English Wikipedia in a single domain [11,51,55,84,91] or more recently in more open-domain scenarios [85].”
NLG Challenge : Neural Models vs . Templates (2018, background): “…can be a syntactic tree (Mani et al., 1999; Barzilay and McKeown, 2005), a set of abstract logic forms (Phillips, 1993), key-value attribute pairs (Chisholm et al., 2017; Dong et al., 2017), vector space embeddings (Dinu and Baroni, 2014), multivariate time series (Sripada et al., 2003), sensor…”
Getting to Production with Few-shot Natural Language Generation Models (2021, methodology): “Also related are earlier approaches that use cycle consistency be-tween parsing and generation models for automatic data cleaning (Nie et al., 2019; Chisholm et al., 2017).”
Variational Template Machine for Data-to-Text Generation (2020, methodology): “There are several recent methods utilizing neural encoder-decoder frameworks to generate text description from data tables (Lebret et al., 2016; Bao et al., 2018; Chisholm et al., 2017; Liu et al., 2018).”
Building Adaptive Acceptability Classifiers for Neural NLG (2021, background): “Also related are earlier approaches that use cycle consistency between parsing and generation models for automatic data cleaning (Nie et al., 2019; Chisholm et al., 2017).”
Describing a Knowledge Base (2018, background): “Several studies (Lebret et al., 2016; Chisholm et al., 2017; Kaffee et al., 2018a,b; Liu et al., 2018; Sha et al., 2018) generate a person’s biography from an input structure, which are closely related to our task.”
Have Your Text and Use It Too! End-to-End Neural Data-to-Text Generation with Semantic Fidelity (2020, background): “Semantic Fidelity Guarantees: To improve semantic fidelity (how accurately the generated text conveys the meaning) in E-to-E architectures, one approach has been to train reverse “Text-to-Data” models (Chisholm et al., 2017; Agarwal et al., 2018).”
GTR-LSTM: A Triple Encoder for Sentence Generation from RDF Data (2018, background): “This task has many applications, such as question answering (Bordes et al., 2014; Fader et al., 2014), proﬁle summarizing (Lebret et al., 2016; Chisholm et al., 2017), and automatic weather forecasting (Mei et al., 2016).”
Generating Live Soccer-Match Commentary from Play Data (2019, background): “Recent work on data-to-text generation includes product review generation (Dong et al. 2017) and biography generation (Lebret, Grangier, and Auli 2016; Liu et al. 2018; Hachey, Radford, and Chisholm 2017; Sha et al. 2018).”
Towards Faithfulness in Open Domain Table-to-text Generation from an Entity-centric View (2021, background): “Recently, end-to-end learning becomes a trend (Mei, Bansal, and Walter 2016; Chisholm, Radford, and Hachey 2017; Kaffee et al. 2018; Jhamtani et al. 2018; Bao et al. 2018; Liu et al. 2019a; Duˇsek, Novikova, and Rieser 2020) in this ﬁeld.”
Operations Guided Neural Networks for High Fidelity Data-To-Text Generation (2018, background): “, 2017), short biographies generation from Wikipedia tables (Lebret et al., 2016; Chisholm et al., 2017) and comments generation based on stock prices (Murakami et al.”
A Sequence-to-Sequence Model for Semantic Role Labeling (2018, methodology): “, 2014; Cho et al., 2014) but soon proved to be useful for any kind of problem that could be represented as a mapping between source and target sequences. Vinyals et al. (2015) demonstrate that constituent parsing can be formulated as a seq2seq problem by linearizing the parse tree.”
Revisiting Challenges in Data-to-Text Generation with Fact Grounding (2020, background): “Data-to-text generation aims at automatically producing descriptive natural language texts to convey the messages embodied in structured data formats, such as database records (Chisholm et al., 2017), knowledge graphs (Gardent et al., 2017a), and tables (Lebret et al., 2016; Wiseman et al., 2017).”
Learning Neural Templates for Text Generation (2018, methodology): “With the continued success of encoder-decoder models for machine translation and related tasks, there has been great interest in extending these methods to build general-purpose, data-driven natural language generation (NLG) systems (Mei et al., 2016; Dušek and Jurcıcek, 2016; Lebret et al., 2016; C”
PaperRobot: Incremental Draft Generation of Scientific Ideas (2019, background): “…et al., 2016; Trisedya et al., 2018; Xu et al., 2018; Madotto et al., 2018; Nie et al., 2018), biographies based on attributes (Lebret et al., 2016; Chisholm et al., 2017; Liu et al., 2018; Sha et al., 2018; Kaffee et al., 2018; Wang et al., 2018a; Wiseman et al., 2018), and image/video captions…”