A list of recent papers about Knowledge-based Machine Reading Comprehension (KMRC).
Contributed by Luxi Xing.
Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China.
We recommend to follow this link (https://github.com/XingLuxi/KMRC-Research-Archive) to get more information.
( we will continuously update this list.)
Content
- Survey Papers
- Benchmark-Datasets
- Cloze-Style-Tasks
- Span-Extraction-Tasks
- Multiple-Choice-Tasks
- Generation-Tasks
Note: papers about KBQA will be not included in this list.
Survey Papers
For more detail, can refer to this link
Neural Machine Reading Comprehension: Methods and Trends. 2019. [paper]
Authors: Shanshan Liu, Xin Zhang, Sheng Zhang, Hui Wang, Weiming Zhang
Machine Reading Comprehension: a Literature Review. 2019. [paper]
Authors: Xin Zhang, An Yang, Sujian Li, Yizhong Wang
Natural Language QA Approaches using Reasoning with External Knowledge. 2020. [paper]
Authors: Chitta Baral, Pratyay Banerjee, Kuntal Pal, Arindam Mitra
Cloze Style Tasks
| Title | Publish | Tasks | Links |
|---|---|---|---|
| Reasoning with Heterogeneous Knowledge for Commonsense Machine Comprehension | ACL 2017 |
SCT | paper |
| World Knowledge for Reading Comprehension: Rare Entity Prediction with Hierarchical LSTMs Using External Descriptions | EMNLP 2017 |
Rare Entity Prediction | paper |
| Knowledgeable Reader: Enhancing Cloze-Style Reading Comprehension with External Commonsense Knowledge | ACL 2018 |
Common Nouns | paper note |
| A Multi-Attention based Neural Network with External Knowledge for Story Ending Predicting Task | COLING 2018 |
SCT | paper |
| Incorporating Structured Commonsense Knowledge in Story Completion | AAAI 2018 |
SCT | paper |
| Story Ending Prediction by Transferable BERT | IJCAI 2019 |
SCT | paper |
| Toward Better Storylines with Sentence-Level Language Models | ACL 2020 |
SCT | paper |
Span Extraction Tasks
| Title | Publish | Tasks | Links |
|---|---|---|---|
| Dynamic Integration of Background Knowledge in Neural NLU Systems | 2018 | SQuAD/ TriviaQA |
paper note |
| Explicit Utilization of General Knowledge in Machine Reading Comprehension | ACL 2019 |
SQuAD | paper note |
| Enhancing Pre-Trained Language Representations with Rich Knowledge for Machine Reading Comprehension | ACL 2019 |
SQuAD/ ReCoRD |
paper note |
| Machine Reading Comprehension Using Structural Knowledge Graph-aware Network | EMNLP 2019 short |
ReCoRD | paper |
+ SG-Net: Syntax-Guided Machine Reading Comprehension |
AAAI 2020 |
SQuAD2.0 RACE |
paper |
+ Semantics-aware BERT for Language Understanding |
AAAI 2020 |
SQuAD2.0 | paper |
+ Entities as Experts: Sparse Memory Access with Entity Supervision |
2020 | TriviaQA | paper |
+ Semantics-Aware Inferential Network for Natural Language Understanding |
2020 | * | paper |
+indicates works regarding injecting knowledge to improve performance on the datasets not included in this summary.
Multiple Choice Tasks
| Title | Publish | Tasks | Links |
|---|---|---|---|
| Yuanfudao at SemEval-2018 Task 11: Three-way Attention and Relational Knowledge for Commonsense Machine Comprehension | SemEval 2018 |
SemEval-2018 Task 11 | paper code |
| Improving Question Answering by Commonsense-Based Pre-Training | AAAI 2019 |
ARC/ OpenBookQA/ SemEval-2018 Task 11 |
paper |
| Improving Machine Reading Comprehension with General Reading Strategies | NAACL 2019 |
ARC/ OpenBookQA/ MCTest/ SemEval-2018 Task 11/ SCT/ MultiRC |
paper code |
| Ranking and Selecting Multi-Hop Knowledge Paths to Better Predict Human Needs | NAACL 2019 |
story commonsense | paper code |
| Incorporating Relation Knowledge into Commonsense Reading Comprehension with Multi-task Learning | CIKM 2019 |
SemEval-2018 Task 11 / SCT | paper |
| Explain Yourself! Leveraging Language Models for Commonsense Reasoning | ACL 2019 |
CommonsenseQA | paper code note |
| Careful Selection of Knowledge to solve Open Book Question Answering | ACL 2019 |
OpenBookQA | paper |
| Improving Question Answering with External Knowledge | EMNLP MRQA 2019 |
ARC/ OpenBookQA |
paper note |
| KagNet: Knowledge-Aware Graph Networks for Commonsense Reasoning | EMNLP 2019 |
CommonsenseQA | paper code note |
| What’s Missing: A Knowledge Gap Guided Approach for Multi-hop Question Answering | EMNLP 2019 |
OpenBookQA | paper note |
| BIG MOOD: Relating Transformers to Explicit Commonsense Knowledge | EMNLP COIN 2019 |
MCScripts v2 | paper |
| Align, Mask and Select: A Simple Method for Incorporating Commonsense Knowledge into Language Representation Models | 2019 | CSQA WSC |
paper |
| Abductive Reasoning as Self-Supervision for Common Sense Question Answering | 2019 | Swag HellaSwag |
paper |
| Exploring ways to incorporate additional knowledge to improve Natural Language Commonsense Question Answering | 2019 | ANLI/ SocialIQA |
paper note |
| Graph-Based Reasoning over Heterogeneous External Knowledge for Commonsense Question Answering | AAAI 2020 |
CSQA | paper note |
| Dynamic Knowledge Graph Construction for Zero-shot Commonsense Question Answering | 2019 | SIQA | paper note |
| K-ADAPTER: Infusing Knowledge into Pre-Trained Models with Adapters | 2020 | CosmosQA | paper note |
| Unsupervised Commonsense Question Answering with Self-Talk | 2020 | CSQA SIQA |
paper note |
| L2R2: Leveraging Ranking for Abductive Reasoning | SIGIR 2020 |
ANLI | paper |
| Unsupervised Alignment-based Iterative Evidence Retrieval for Multi-hop Question Answering | ACL 2020 |
MultiRC QASC |
paper |
| Logic-Guided Data Augmentation and Regularization for Consistent Question Answering | ACL 2020 |
WIQA QuaREL |
paper |
| Connecting the Dots: A Knowledgeable Path Generator for Commonsense Question Answering | 2020 | CSQA OBQA |
paper |
| Fusing Context Into Knowledge Graph for Commonsense Reasoning | 2020 | CSQA | paper |
Generation Tasks
Also known as Free-form Answer Tasks
| Title | Publish | Tasks | Links |
|---|---|---|---|
| Commonsense for Generative Multi-Hop Question Answering Tasks | EMNLP 2018 |
NarrativeQA/ WikiHop |
paper code note |
| COMET: Commonsense Transformers for Automatic Knowledge Graph Construction | ACL 2019 |
Atomic | paper code note |
| Incorporating External Knowledge into Machine Reading for Generative Question Answering | EMNLP 2019 |
MS MARCO | paper note |
| Modeling Event Background for If-Then Commonsense Reasoning Using Context-aware Variational Autoencoder | EMNLP 2019 |
Event2Mind | paper |
| Dynamic Knowledge Graph Construction for Zero-shot Commonsense Question Answering | 2019 | SocialIQA StoryCommonsense |
paper |
| Injecting Numerical Reasoning Skills into Language Models | ACL 2020 |
DROP | paper |
Benchmark Datasets
[COPA] Choice of Plausible Alternatives: An Evaluation of Commonsense Causal Reasoning. AAAI,2011. [paper / data]
Authors: Melissa Roemmele, Cosmin Adrian Bejan, Andrew S. Gordon
- Type: Multiple-Choice;
[WSC] The Winograd Schema Challenge. AAAI,2011. [paper /data]
Authors: Hector J. Levesque, Ernest Davis, Leora Morgenstern
- Type: Multiple-Choice;
[ROCStories; SCT] A Corpus and Cloze Evaluation for Deeper Understanding of Commonsense Stories. NAACL,2016. [paper / data]
Authors: Nasrin Mostafazadeh, Nathanael Chambers, Xiaodong He, Devi Parikh, Dhruv Batra, Lucy Vanderwende, Pushmeet Kohli, James Allen
- Type: Cloze;
[NarrativeQA] The NarrativeQA Reading Comprehension Challenge. TACL,2018. [paper / data]
Authors: Tomáš Kočiský, Jonathan Schwarz, Phil Blunsom, Chris Dyer, Karl Moritz Hermann, Gábor Melis, Edward Grefenstette
- Type: Generation;
[SemEval-2018 Task 11] MCScript: A Novel Dataset for Assessing Machine Comprehension Using Script Knowledge. LERC,2018. [paper / data]
Authors: Simon Ostermann, Ashutosh Modi, Michael Roth, Stefan Thater, Manfred Pinkal
- Type: Multiple-Choice;
[story-commonsense] Modeling Naive Psychology of Characters in Simple Commonsense Stories. ACL,2018. [paper / data]
Authors: Hannah Rashkin, Antoine Bosselut, Maarten Sap, Kevin Knight, Yejin Choi
- Type: Multiple-Choice;
Event2Mind: Commonsense Inference on Events, Intents, and Reactions. ACL,2018. [paper / data]
Authors: Hannah Rashkin, Maarten Sap, Emily Allaway, Noah A. Smith, Yejin Choi
- Types: Generation;
ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning. AAAI,2019. [paper / data]
Authors: Maarten Sap, Ronan LeBras, Emily Allaway, Chandra Bhagavatula, Nicholas Lourie, Hannah Rashkin, Brendan Roof, Noah A. Smith, Yejin Choi
- Types: Generation;
[ARC] Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge. 2018. [paper / data]
Authors: Peter Clark, Isaac Cowhey, Oren Etzioni, Tushar Khot, Ashish Sabharwal, Carissa Schoenick, Oyvind Tafjord
- Type: Multiple-Choice;
[OpenBookQA] Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering. EMNLP,2018. [paper / data]
Authors: Todor Mihaylov, Peter Clark, Tushar Khot, Ashish Sabharwal
- Type: Multiple-Choice;
ReCoRD: Bridging the Gap between Human and Machine Commonsense Reading Comprehension. 2018. [paper / data]
Authors: Sheng Zhang, Xiaodong Liu, Jingjing Liu, Jianfeng Gao, Kevin Duh, Benjamin Van Durme
- Type: Cloze;
CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge. NAACL,2019. [paper / data]
Authors: Alon Talmor, Jonathan Herzig, Nicholas Lourie, Jonathan Berant
- Type: Multiple-Choice;
ChID: A Large-scale Chinese IDiom Dataset for Cloze Test. ACL,2019. [paper / data]
Authors: Chujie Zheng, Minlie Huang, Aixin Sun
- Type: Cloze;
[sense-making] Does it Make Sense? And Why? A Pilot Study for Sense Making and Explanation. ACL,2019. [paper / data]
Authors: Cunxiang Wang, Shuailong Liang, Yue Zhang, Xiaonan Li, Tian Gao
- Type: Multiple-Choice;
HellaSwag: Can a Machine Really Finish Your Sentence? ACL,2019. [paper / data]
Authors: Rowan Zellers, Ari Holtzman, Yonatan Bisk, Ali Farhadi, Yejin Choi
- Type: Multiple-Choice;
SocialIQA: Commonsense Reasoning about Social Interactions. EMNLP,2019. [paper / data]
Authors: Maarten Sap, Hannah Rashkin, Derek Chen, Ronan LeBras, Yejin Choi
- Type: Multiple-Choice;
[ANLI] Abductive Commonsense Reasoning. 2019. [paper / data]
Authors: Chandra Bhagavatula, Ronan Le Bras, Chaitanya Malaviya, Keisuke Sakaguchi, Ari Holtzman, Hannah Rashkin, Doug Downey, Scott Wen-tau Yih, Yejin Choi
- Type: Multiple-Choice;
Cosmos QA: Machine Reading Comprehension with Contextual Commonsense Reasoning. EMNLP,2019. [paper / data]
Authors: Lifu Huang, Ronan Le Bras, Chandra Bhagavatula, Yejin Choi
- Type: Multiple-Choice;
CODAH: An Adversarially-Authored Question Answering Dataset for Common Sense. ACL,2019,workshop. [paper / data ]
Authors: Michael Chen, Mike D’Arcy, Alisa Liu, Jared Fernandez, Doug Downey
- Type: Multiple-Choice;
CommonGen: A Constrained Text Generation Dataset Towards Generative Commonsense Reasoning. 2019. [ paper / data ]
Authors: Bill Yuchen Lin, Ming Shen, Yu Xing, Pei Zhou, Xiang Ren
- Type: Generative;
QASC: A Dataset for Question Answering via Sentence Composition. 2019. [paper / data ]
Authors: Tushar Khot, Peter Clark, Michal Guerquin, Peter Jansen, Ashish Sabharwal
DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs. NAACL,2019. [paper]
Authros: Dheeru Dua, Yizhong Wang, Pradeep Dasigi, Gabriel Stanovsky, Sameer Singh, Matt Gardner
Note: Only consider the benchmark datasets/tasks which require knowledge to complete.