Data-to-Text-Generation

[TOC]

News:

2021.09.27

• Papers
  1. DART: Open-Domain Structured Data Record to Text Generation NAACL2021
  2. Promoting Graph Awareness in Linearized Graph-to-Text Generation ACL2021 Findings
  3. De-Confounded Variational Encoder-Decoder for Logical Table-to-Text Generation ACL2021
  4. Towards Table-to-Text Generation with Numerical Reasoning ACL2021
  5. Improving Encoder by Auxiliary Supervision Tasks for Table-to-Text Generation ACL2021
  6. WIKITABLET: A Large-Scale Data-to-Text Dataset for Generating Wikipedia Article Sections ACL2021 Findings
  7. Structural Adapters in Pretrained Language Models for AMR-to-text Generation EMNLP 2021
  8. Plan-then-Generate: Controlled Data-to-Text Generation via Planning EMNLP2021 Findings
  9. Few-Shot Table-to-Text Generation with Prototype Memory EMNLP2021 Findings
  10. Smelting Gold and Silver for Improved Multilingual AMR-to-Text Generation EMNLP2021
• Datasetes
  1. DART: Open-Domain Structured Data Record to Text Generation NAACL2021
  2. Towards Table-to-Text Generation with Numerical Reasoning ACL2021
  3. WIKITABLET: A Large-Scale Data-to-Text Dataset for Generating Wikipedia Article Sections ACL2021 Findings

2021.07.21

• Papers
  1. WikiGraphs: A Wikipedia Text - Knowledge Graph Paired Dataset NAACL|TextGraphs2021
  2. Data-to-text Generation with Macro Planning TACL2021
  3. Modeling Graph Structure via Relative Position for Text Generation from Knowledge Graphs NAACL|TextGraphs2021
  4. Generating Landmark Navigation Instructions from Maps as a Graph-to-Text Problem ACL2021
• Datasets
  1. WikiGraphs NAACL|TextGraphs2021
  2. Map2Seq ACL2021

Content

1. Papers
2. Datasets
3. Evaluation Metrics

1. Papers

2016

1. Neural Text Generation from Structured Data with Application to the Biography Domain EMNLP2016
   • Code:Official

2017

1. Challenges in Data-to-Document Generation EMNLP2017
   • Code: Official
2. Order-planning neural text generation from structured data AAAI2018
3. Table-to-text Generation by Structure-aware Seq2seq Learning AAAI2018
   • Code:Official
4. Table-to-Text: Describing Table Region with Natural Language AAAI2018
5. A Graph-to-Sequence Model for AMR-to-Text Generation ACL2018
   • Code: Official
6. Graph-to-Sequence Learning using Gated Graph Neural Networks ACL2018
   • Code: Official
7. Generating Descriptions from Structured Data Using a Bifocal Attention Mechanism and Gated Orthogonalization NAACL2018
   • Code: Official
8. A mixed hierarchical attention based encoder-decoder approach for standard summarizaion NAACL2018

2018

1. Operation-guided Neural Networks for High Fidelity Data-To-Text Generation EMNLP2018
2. Learning Neural Templates for Text Generation EMNLP2018
   • Code: Official
3. Learning Latent Semantic Annotations for Grounding Natural Language to Structured Data EMNLP2018
   • Code: Official
4. Data2Text Studio: Automated Text Generation from Structured Data EMNLP2018
5. Data-to-Text Generation with Content Selection and Planning AAAI2019
   • Code: Official
6. Hierarchical Encoder with Auxiliary Supervision for Neural Table-to-Text Generation: Learning Better Representation for Tables AAAI2019
7. Key Fact as Pivot: A Two-Stage Model for Low Resource Table-to-Text Generation ACL2019
8. Learning to Select, Track, and Generate for Data-to-Text ACL2019
   • Code: Official
9. Towards Comprehensive Description Generation from Factual Attribute-value Tables ACL2019
10. Data-to-text Generation with Entity Modeling ACL2019
    • Code: Official
11. Handling Divergent Reference Texts when Evaluating Table-to-Text Generation ACL2019
    • Code: Official
12. Step-by-Step: Separating Planning from Realization in Neural Data-to-Text Generation NAACL2019
    • Code: Official
13. Text Generation from Knowledge Graphs with Graph Transformers NAACL2019
    • Code: Official
14. Structural Neural Encoders for AMR-to-text Generation NAACL2019 NAACL2019
    • Code: Official
15. Deep Graph Convolutional Encoders for Structured Data to Text Generation INLG2018
    • Code: Official
16. Densely Connected Graph Convolutional Networks for Graph-to-Sequence Learning TACL2018
    • Code: Official
17. ...

2019

1. Enhancing Neural Data-To-Text Generation Models with External Background Knowledge EMNLP2019
   • Code: Official
2. Neural data-to-text generation: A comparison between pipeline and end-to-end architectures EMNLP2019
   • Code: Official
3. Table-to-Text Generation with Effective Hierarchical Encoder on Three dimensions (Row, Column and Time) EMNLP2019
   • Code: Official
4. Modeling Graph Structure in Transformer for Better AMR-to-Text Generation EMNLP2019
   • Code: Official
5. Enhanced Transformer Model for Data-to-Text Generation EMLP-WGNT2019
   • Code: Official
6. Selecting, Planning, and Rewriting: A Modular Approach for Data-to-Document Generation and Translation EMNLP2019-short
7. Long and Diverse Text Generation with Planning-based Hierarchical Variational Model EMNLP2019
   • Code: Official
8. Enhancing AMR-to-Text Generation with Dual Graph Representations EMNLP2019
   • Code: Official
9. An Encoder with non-Sequential Dependency for Neural Data-to-Text Generation INLG2019
10. Controlling Contents in Data-to-Document Generation with Human-Designed Topic Labels INLG2019
11. Revisiting Challenges in Data-to-Text Generation with Fact Grounding INLG2019
    • Code: Official
12. Graph Transformer for Graph-to-Sequence Learning AAAI2020
    • Code: Official
13. Sentence Generation for Entity Description with Content-plan Attention AAAI2020
    • Code: Official
14. Learning to Select Bi-Aspect Information for Document-Scale Text Content Manipulation AAAI2020
    • Code: Official
15. Variational Template Machine for Data-to-Text Generation ICLR2020
16. Towards Faithful Neural Table-to-Text Generation with Content-Matching Constraints ACL2020
17. Neural Data-to-Text Generation via Jointly Learning the Segmentation and Correspondence ACL2020
18. Bridging the Structural Gap Between Encoding and Decoding for Data-To-Text Generation ACL2020
    • Code: Official
19. Heterogeneous Graph Transformer for Graph-to-Sequence Learning ACL2020
    • Code: Official
20. Structural Information Preserving for Graph-to-Text Generation ACL2020
    • Code: Official
21. Line Graph Enhanced AMR-to-Text Generation with Mix-Order Graph Attention Networks ACL2020
    • Code: Official
22. GPT-too: A Language-Model-First Approach for AMR-to-Text Generation ACL2020
    • Code: Official
23. Logical Natural Language Generation from Open-Domain Tables ACL2020
    • Code: Official
24. A Generative Model for Joint Natural Language Understanding and Generation ACL2020
    • Code: Official
25. Two Birds, One Stone: A Simple, Unified Model for Text Generation from Structured and Unstructured Data ACL2020
    • Code: Official
26. Infobox-to-text Generation with Tree-like PLanning based Attention Network IJCAI2020
27. Better AMR-To-Text Generation with Graph Structure Reconstruction IJCAI2020
    • Code: Official
28. RDF-to-Text Generation with Graph-augmented Structural Neural Encoders IJCAI2020
    • Code: Official
29. A Hierarchical Model for Data-to-Text Generation ECIR2020
    • Code: Official
30. ...

2020

1. ToTTo: A Controlled Table-To-Text Generation Dataset EMNLP2020
   • Code: Official
2. CycleGT: Unsupervised Graph-to-Text and Text-to-Graph Generation via Cycle Training NIPS2020
   • Code: Official
3. Modeling Global and Local Node Contexts for Text Generation from Knowledge Graphs TACL2020
   • Code: Official
4. AMR-to-text Generation with Graph Transformer TACL2020
   • Code: Official
5. Have Your Text and Use It Too! End-to-End Neural Data-to-Text Generation with Semantic Fidelity COLING2020
   • Code: Official
6. Investigating Pretrained Language Models for Graph-to-Text Generation arXiv2020
   • Code: Official
7. Logic2Text: High-Fidelity Natural Language Generation from Logical Forms EMNLP2020
   • Code: Official
8. KGPT: Knowledge-Grounded Pre-Training for Data-to-Text Generation EMNLP2020
   • Code: Official
9. Online Back-Parsing for AMR-to-Text Generation EMNLP2020
   • Code: Official
10. Lightweight, Dynamic Graph Convolutional Networks for AMR-to-Text Generation EMNLP2020
    • Code: Official
11. Stepwise Extractive Summarization and Planning with Structured Transformers EMNLP2020
    • Code: Official
12. Partially-Aligned Data-to-Text Generation with Distant Supervision EMNLP2020
    • Code: Official
13. GenWiki: A Dataset of 1.3 Million Content-Sharing Text and Graphs for Unsupervised Graph-to-Text Generation COLING2020
    • Code: Official
14. Enhancing Content Planning for Table-to-Text Generation with Data Understanding and Verification EMNLP2020 Findings
    • Code: Official
15. Multilingual AMR-to-Text Generation EMNLP2020
    • Code: Official
16. ENT-DESC: Entity Description Generation by Exploring Knowledge Graph EMNLP2020
    • Code: Official
17. An Unsupervised Joint System for Text Generation from Knowledge Graphs and Semantic Parsing EMNLP2020
    • Code: Official
18. Make Templates Smarter: A Template Based Data2Text System Powered by Text Stitch Model EMNLP2020 Findings
19. TableGPT: Few-shot Table-to-Text Generation with Table Structure Reconstruction and Content Matching COLING2020
20. Towards Faithfulness in Open Domain Table-to-text Generation from an Entity-centric View AAAI2021
21. Neural Data-to-Text Generation with LM-based Text Augmentation EACL2021
22. DART: Open-Domain Structured Data Record to Text Generation NAACL2021
    • Code: Official
23. Modeling Graph Structure via Relative Position for Text Generation from Knowledge Graphs NAACL|TextGraphs2021
    • Code: Official
24. WikiGraphs: A Wikipedia Text - Knowledge Graph Paired Dataset NAACL|TextGraphs2021
    • Code: Official
25. Data-to-text Generation with Macro Planning TACL2021
    • Code: Official
26. Few-shot Knowledge Graph-to-Text Generation with Pretrained Language Models ACL2021 Findings
    • Code: Official
27. Stage-wise Fine-tuning for Graph-to-Text Generation ACL2021 Workshop
    • Code: Official
28. Generating Landmark Navigation Instructions from Maps as a Graph-to-Text Problem ACL2021
29. Sketch and Refine: Towards Faithful and Informative Table-to-Text Generation ACL2021 Findings
30. Promoting Graph Awareness in Linearized Graph-to-Text Generation ACL2021 Findings
31. De-Confounded Variational Encoder-Decoder for Logical Table-to-Text Generation ACL2021
32. Towards Table-to-Text Generation with Numerical Reasoning ACL2021
    • Code: Official
33. Improving Encoder by Auxiliary Supervision Tasks for Table-to-Text Generation ACL2021
    • Code: Official
34. WIKITABLET: A Large-Scale Data-to-Text Dataset for Generating Wikipedia Article Sections ACL2021 Findings
    • Code: Official
35. ...

2021

1. Structural Adapters in Pretrained Language Models for AMR-to-text Generation EMNLP 2021
   • Code: Official
2. Learning to Reason for Text Generation from Scientific Tables ArXiv2021
   • Code: Official
3. Plan-then-Generate: Controlled Data-to-Text Generation via Planning EMNLP2021 Findings
   • Code: Official
4. Few-Shot Table-to-Text Generation with Prototype Memory EMNLP2021Findings
   • Code: Official
5. Smelting Gold and Silver for Improved Multilingual AMR-to-Text Generation EMNLP2021
   • Code: Official
6. ...

2. DataSets

The citation information was updated on Jan 4, 2021

2.1 Table2Text

No	Dataset	Domain	Source	Train/Dev/Test	Cited
1	WIKIBIO	Wikipedia	Neural text generation from structured data with application to the biography domain EMNLP2016	582,695/72,831/72,831	212	Official
2	E2E	Restaurants	The E2E dataset: New challenges for end-to- end generation SIGDIAL2017	4,2061/4,672/4,693	135	Official
3	ROTOWIRE	Basketball	Challenges in Data-to-Document Generation EMNLP2017	3,371/727/728	227	Official
4	WEATHERGOV	Weather	Learning Semantic Correspondences with Less Supervision ACL2009	25,000/1,000/3,528	255	Official
6	ESPN	Basketball	Operation-guided Neural Networks for High Fidelity Data-To-Text Generation EMNLP2018	12,043/1,505/1,506	25	Official
7	ROTOWIRE-MODIFIED	Basketball	Learning to Select, Track, and Generate for Data-to-Text ACL2019	2,705/532/497	14	Official
8	MLB	Baseball	Data-to-text Generation with Entity Modeling ACL2019	22,821/1,739/1,744	27	Official
9	WikiBio (21 domains)	Wikipedia	Enhancing Neural Data-To-Text Generation Models with External Background Knowledge EMNLP2019	-	9	Official
10	Rotowire-FG	Basketball	Revisiting Challenges in Data-to-Text Generation with Fact Grounding INLG2019	5,232/1,125/1,119	5	Official
11	Wikiperson	Wikipedia	Describing a Knowledge Base INLG2018	250,186/30,487/29,982	21	Official
12	LOGICNLG	Wikipedia	Logical Natural Language Generation from Open-Domain Tables ACL2020	28,450/4,260/4,305	12	Official
13	ToTTo	Wikipedia	ToTTo: A Controlled Table-To-Text Generation Dataset EMNLP2020	120,761/7,700/7,700	13	Official
14	SciGen	ArXiv	Learning to Reason for Text Generation from Scientific Tables ArXiv2021	-	-	Official
15	numericNLG	Scientific	Towards Table-to-Text Generation with Numerical Reasoning ACL2021	-	-	Official
16	WIKITABLET	Wikipedia	WIKITABLET: A Large-Scale Data-to-Text Dataset for Generating Wikipedia Article Sections ACL2021 Findings	-/4533/4351	-	Official

2.2 Graph2Text

No	Dataset	Type	Source	Train/Dev/Test	Cited
1	WebNLG	RDF	-	-	-	Official
2	LDC2015E86	AMR	-	16,833/1,368/1,371	-	Official
3	LDC2017T10	AMR	-	36,521/1,368/1,371	-	Official
4	LDC2020T02	AMR	-	55,635/1,722/1,898	-	Official
5	AGENDA	Knowledge Graphs	Text Generation from Knowledge Graphs with Graph Transformers NAACL2019	38,720/1,000/1,000	74	Official
6	LOGIC2TEXT	Wikipedia	Logic2Text: High-Fidelity Natural Language Generation from Logical Forms EMNLP2020	8,566/1,095/1,092	1	Official
7	WITA	Wikipedia	Partially-Aligned Data-to-Text Generation with Distant Supervision EMNLP2020	50,000/5,000/400	2	Official
8	GenWiki	-	GenWiki: A Dataset of 1.3 Million Content-Sharing Text and Graphs for Unsupervised Graph-to-Text Generation COLING2020	-	-	Official
9	ENT-DESC	-	ENT-DESC: Entity Description Generation by Exploring Knowledge Graph EMNLP2020	-	-	Official
10	WikiGraphs	Wikipedia	WikiGraphs: A Wikipedia Text - Knowledge Graph Paired Dataset NAACL|TextGraphs2021	23,431/48/43	-	Official
11	Map2Seq	OpenStreetMap	Generating Landmark Navigation Instructions from Maps as a Graph-to-Text Problem ACL2021	-	-	Official
12	DART	Wikipedia+Restaurant	DART: Open-Domain Structured Data Record to Text Generation NAACL2021	62,659/6,980/12,552	11	Official

3. Evaluation Metrics

No	Metric	Source	Cited
1	BLEU	Bleu: a Method for Automatic Evaluation of Machine Translation ACL2020	14039	-
2	CS, RG, CO	Challenges in Data-to-Document Generation EMNLP2017	227	Official
3	PARENT	Handling Divergent Reference Texts when Evaluating Table-to-Text Generation ACL2019	18	Official

Updating ......