Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Kewei Tu

Kewei Tu

Dynamic Programming in Rank Space: Scaling Structured Inference with Low-Rank HMMs and PCFGs

May 01, 2022
Songlin Yang, Wei Liu, Kewei Tu

Figure 1 for Dynamic Programming in Rank Space: Scaling Structured Inference with Low-Rank HMMs and PCFGs

Figure 2 for Dynamic Programming in Rank Space: Scaling Structured Inference with Low-Rank HMMs and PCFGs

Figure 3 for Dynamic Programming in Rank Space: Scaling Structured Inference with Low-Rank HMMs and PCFGs

Figure 4 for Dynamic Programming in Rank Space: Scaling Structured Inference with Low-Rank HMMs and PCFGs

* Accepted to NAACL 2022

Access Paper or Ask Questions

Modeling Label Correlations for Second-Order Semantic Dependency Parsing with Mean-Field Inference

Apr 07, 2022
Songlin Yang, Kewei Tu

Figure 1 for Modeling Label Correlations for Second-Order Semantic Dependency Parsing with Mean-Field Inference

Figure 2 for Modeling Label Correlations for Second-Order Semantic Dependency Parsing with Mean-Field Inference

Figure 3 for Modeling Label Correlations for Second-Order Semantic Dependency Parsing with Mean-Field Inference

Figure 4 for Modeling Label Correlations for Second-Order Semantic Dependency Parsing with Mean-Field Inference

* 12 pages

Access Paper or Ask Questions

Nested Named Entity Recognition as Latent Lexicalized Constituency Parsing

Mar 09, 2022
Chao Lou, Songlin Yang, Kewei Tu

Figure 1 for Nested Named Entity Recognition as Latent Lexicalized Constituency Parsing

Figure 2 for Nested Named Entity Recognition as Latent Lexicalized Constituency Parsing

Figure 3 for Nested Named Entity Recognition as Latent Lexicalized Constituency Parsing

Figure 4 for Nested Named Entity Recognition as Latent Lexicalized Constituency Parsing

* ACL 2022 camera ready

Access Paper or Ask Questions

DAMO-NLP at SemEval-2022 Task 11: A Knowledge-based System for Multilingual Named Entity Recognition

Mar 01, 2022
Xinyu Wang, Yongliang Shen, Jiong Cai, Tao Wang, Xiaobin Wang, Pengjun Xie, Fei Huang, Weiming Lu, Yueting Zhuang, Kewei Tu, Wei Lu, Yong Jiang

Figure 1 for DAMO-NLP at SemEval-2022 Task 11: A Knowledge-based System for Multilingual Named Entity Recognition

Figure 2 for DAMO-NLP at SemEval-2022 Task 11: A Knowledge-based System for Multilingual Named Entity Recognition

Figure 3 for DAMO-NLP at SemEval-2022 Task 11: A Knowledge-based System for Multilingual Named Entity Recognition

Figure 4 for DAMO-NLP at SemEval-2022 Task 11: A Knowledge-based System for Multilingual Named Entity Recognition

* Our Knowledge-based NER system wins 10 out of 13 tracks in the SemEval-2022 MultiCoNER shared task

Access Paper or Ask Questions

Bottom-Up Constituency Parsing and Nested Named Entity Recognition with Pointer Networks

Oct 11, 2021
Songlin Yang, Kewei Tu

Figure 1 for Bottom-Up Constituency Parsing and Nested Named Entity Recognition with Pointer Networks

Figure 2 for Bottom-Up Constituency Parsing and Nested Named Entity Recognition with Pointer Networks

Figure 3 for Bottom-Up Constituency Parsing and Nested Named Entity Recognition with Pointer Networks

Figure 4 for Bottom-Up Constituency Parsing and Nested Named Entity Recognition with Pointer Networks

Access Paper or Ask Questions

Combining (second-order) graph-based and headed span-based projective dependency parsing

Aug 12, 2021
Songlin Yang, Kewei Tu

Figure 1 for Combining (second-order) graph-based and headed span-based projective dependency parsing

Figure 2 for Combining (second-order) graph-based and headed span-based projective dependency parsing

Figure 3 for Combining (second-order) graph-based and headed span-based projective dependency parsing

Figure 4 for Combining (second-order) graph-based and headed span-based projective dependency parsing

* 5 pages

Access Paper or Ask Questions

Headed Span-Based Projective Dependency Parsing

Aug 10, 2021
Songlin Yang, Kewei Tu

Figure 1 for Headed Span-Based Projective Dependency Parsing

Figure 2 for Headed Span-Based Projective Dependency Parsing

Figure 3 for Headed Span-Based Projective Dependency Parsing

Figure 4 for Headed Span-Based Projective Dependency Parsing

Access Paper or Ask Questions

Enhanced Universal Dependency Parsing with Automated Concatenation of Embeddings

Jul 06, 2021
Xinyu Wang, Zixia Jia, Yong Jiang, Kewei Tu

Figure 1 for Enhanced Universal Dependency Parsing with Automated Concatenation of Embeddings

Figure 2 for Enhanced Universal Dependency Parsing with Automated Concatenation of Embeddings

Figure 3 for Enhanced Universal Dependency Parsing with Automated Concatenation of Embeddings

Figure 4 for Enhanced Universal Dependency Parsing with Automated Concatenation of Embeddings

* Second Place in IWPT 2021 shared task, 7 pages

Access Paper or Ask Questions

Improving Named Entity Recognition by External Context Retrieving and Cooperative Learning

Jun 02, 2021
Xinyu Wang, Yong Jiang, Nguyen Bach, Tao Wang, Zhongqiang Huang, Fei Huang, Kewei Tu

Figure 1 for Improving Named Entity Recognition by External Context Retrieving and Cooperative Learning

Figure 2 for Improving Named Entity Recognition by External Context Retrieving and Cooperative Learning

Figure 3 for Improving Named Entity Recognition by External Context Retrieving and Cooperative Learning

Figure 4 for Improving Named Entity Recognition by External Context Retrieving and Cooperative Learning

* Accepted to ACL 2021, 12 pages

Access Paper or Ask Questions

Neural Bi-Lexicalized PCFG Induction

May 31, 2021
Songlin Yang, Yanpeng Zhao, Kewei Tu

Figure 1 for Neural Bi-Lexicalized PCFG Induction

Figure 2 for Neural Bi-Lexicalized PCFG Induction

Figure 3 for Neural Bi-Lexicalized PCFG Induction

Figure 4 for Neural Bi-Lexicalized PCFG Induction

* To appear in ACL 2021 main conference

Access Paper or Ask Questions

PCFGs Can Do Better: Inducing Probabilistic Context-Free Grammars with Many Symbols

Apr 28, 2021
Songlin Yang, Yanpeng Zhao, Kewei Tu

Figure 1 for PCFGs Can Do Better: Inducing Probabilistic Context-Free Grammars with Many Symbols

Figure 2 for PCFGs Can Do Better: Inducing Probabilistic Context-Free Grammars with Many Symbols

Figure 3 for PCFGs Can Do Better: Inducing Probabilistic Context-Free Grammars with Many Symbols

Figure 4 for PCFGs Can Do Better: Inducing Probabilistic Context-Free Grammars with Many Symbols

* NAACL 2021

Access Paper or Ask Questions

Constrained Text Generation with Global Guidance -- Case Study on CommonGen

Mar 12, 2021
Yixian Liu, Liwen Zhang, Wenjuan Han, Yue Zhang, Kewei Tu

Figure 1 for Constrained Text Generation with Global Guidance -- Case Study on CommonGen

Figure 2 for Constrained Text Generation with Global Guidance -- Case Study on CommonGen

Figure 3 for Constrained Text Generation with Global Guidance -- Case Study on CommonGen

Figure 4 for Constrained Text Generation with Global Guidance -- Case Study on CommonGen

Access Paper or Ask Questions

Neural Latent Dependency Model for Sequence Labeling

Nov 10, 2020
Yang Zhou, Yong Jiang, Zechuan Hu, Kewei Tu

Figure 1 for Neural Latent Dependency Model for Sequence Labeling

Figure 2 for Neural Latent Dependency Model for Sequence Labeling

Figure 3 for Neural Latent Dependency Model for Sequence Labeling

Figure 4 for Neural Latent Dependency Model for Sequence Labeling

Access Paper or Ask Questions

Second-Order Unsupervised Neural Dependency Parsing

Oct 28, 2020
Songlin Yang, Yong Jiang, Wenjuan Han, Kewei Tu

Figure 1 for Second-Order Unsupervised Neural Dependency Parsing

Figure 2 for Second-Order Unsupervised Neural Dependency Parsing

Figure 3 for Second-Order Unsupervised Neural Dependency Parsing

Figure 4 for Second-Order Unsupervised Neural Dependency Parsing

* COLING 2020 camera ready

Access Paper or Ask Questions

Adversarial Attack and Defense of Structured Prediction Models

Oct 18, 2020
Wenjuan Han, Liwen Zhang, Yong Jiang, Kewei Tu

Figure 1 for Adversarial Attack and Defense of Structured Prediction Models

Figure 2 for Adversarial Attack and Defense of Structured Prediction Models

Figure 3 for Adversarial Attack and Defense of Structured Prediction Models

Figure 4 for Adversarial Attack and Defense of Structured Prediction Models

* Accepted to EMNLP 2020

Access Paper or Ask Questions

AIN: Fast and Accurate Sequence Labeling with Approximate Inference Network

Oct 12, 2020
Xinyu Wang, Yong Jiang, Nguyen Bach, Tao Wang, Zhongqiang Huang, Fei Huang, Kewei Tu

Figure 1 for AIN: Fast and Accurate Sequence Labeling with Approximate Inference Network

Figure 2 for AIN: Fast and Accurate Sequence Labeling with Approximate Inference Network

Figure 3 for AIN: Fast and Accurate Sequence Labeling with Approximate Inference Network

Figure 4 for AIN: Fast and Accurate Sequence Labeling with Approximate Inference Network

* Accept to Main Conference of EMNLP 2020 (Short). Camera-ready, 8 Pages

Access Paper or Ask Questions

Structural Knowledge Distillation

Oct 10, 2020
Xinyu Wang, Yong Jiang, Zhaohui Yan, Zixia Jia, Nguyen Bach, Tao Wang, Zhongqiang Huang, Fei Huang, Kewei Tu

Figure 1 for Structural Knowledge Distillation

Figure 2 for Structural Knowledge Distillation

Figure 3 for Structural Knowledge Distillation

Figure 4 for Structural Knowledge Distillation

* Under review as a conference paper of ICLR 2021. 15 pages

Access Paper or Ask Questions

Automated Concatenation of Embeddings for Structured Prediction

Oct 10, 2020
Xinyu Wang, Yong Jiang, Nguyen Bach, Tao Wang, Zhongqiang Huang, Fei Huang, Kewei Tu

Figure 1 for Automated Concatenation of Embeddings for Structured Prediction

Figure 2 for Automated Concatenation of Embeddings for Structured Prediction

Figure 3 for Automated Concatenation of Embeddings for Structured Prediction

Figure 4 for Automated Concatenation of Embeddings for Structured Prediction

* We propose ACE, which achieves new SOTA for 6 NLP tasks over 23 datasets. Under review as a conference paper at ICLR 2021. 19 pages

Access Paper or Ask Questions

More Embeddings, Better Sequence Labelers?

Oct 10, 2020
Xinyu Wang, Yong Jiang, Nguyen Bach, Tao Wang, Zhongqiang Huang, Fei Huang, Kewei Tu

Figure 1 for More Embeddings, Better Sequence Labelers?

Figure 2 for More Embeddings, Better Sequence Labelers?

Figure 3 for More Embeddings, Better Sequence Labelers?

Figure 4 for More Embeddings, Better Sequence Labelers?

* Accepted to Findings of EMNLP 2020. Camera-ready, 16 pages

Access Paper or Ask Questions

Second-Order Neural Dependency Parsing with Message Passing and End-to-End Training

Oct 10, 2020
Xinyu Wang, Kewei Tu

Figure 1 for Second-Order Neural Dependency Parsing with Message Passing and End-to-End Training

Figure 2 for Second-Order Neural Dependency Parsing with Message Passing and End-to-End Training

Figure 3 for Second-Order Neural Dependency Parsing with Message Passing and End-to-End Training

Figure 4 for Second-Order Neural Dependency Parsing with Message Passing and End-to-End Training

* Accepted to AACL 2020. 7 pages

Access Paper or Ask Questions

A Survey of Unsupervised Dependency Parsing

Oct 04, 2020
Wenjuan Han, Yong Jiang, Hwee Tou Ng, Kewei Tu

Figure 1 for A Survey of Unsupervised Dependency Parsing

Figure 2 for A Survey of Unsupervised Dependency Parsing

* COLING 2020

Access Paper or Ask Questions

Fast and Accurate Sequence Labeling with Approximate Inference Network

Sep 17, 2020
Xinyu Wang, Yong Jiang, Nguyen Bach, Tao Wang, Zhongqiang Huang, Fei Huang, Kewei Tu

Figure 1 for Fast and Accurate Sequence Labeling with Approximate Inference Network

Figure 2 for Fast and Accurate Sequence Labeling with Approximate Inference Network

Figure 3 for Fast and Accurate Sequence Labeling with Approximate Inference Network

Figure 4 for Fast and Accurate Sequence Labeling with Approximate Inference Network

* Accept to Main Conference of EMNLP 2020 (Short). 8 Pages

Access Paper or Ask Questions

Enhanced Universal Dependency Parsing with Second-Order Inference and Mixture of Training Data

Jun 11, 2020
Xinyu Wang, Yong Jiang, Kewei Tu

Figure 1 for Enhanced Universal Dependency Parsing with Second-Order Inference and Mixture of Training Data

Figure 2 for Enhanced Universal Dependency Parsing with Second-Order Inference and Mixture of Training Data

Figure 3 for Enhanced Universal Dependency Parsing with Second-Order Inference and Mixture of Training Data

Figure 4 for Enhanced Universal Dependency Parsing with Second-Order Inference and Mixture of Training Data

* IWPT 2020 shared task. After fixing the bug, our proposed parser performs better than the team that ranked 1st in the official results

Access Paper or Ask Questions

Structure-Level Knowledge Distillation For Multilingual Sequence Labeling

Apr 29, 2020
Xinyu Wang, Yong Jiang, Nguyen Bach, Tao Wang, Fei Huang, Kewei Tu

Figure 1 for Structure-Level Knowledge Distillation For Multilingual Sequence Labeling

Figure 2 for Structure-Level Knowledge Distillation For Multilingual Sequence Labeling

Figure 3 for Structure-Level Knowledge Distillation For Multilingual Sequence Labeling

Figure 4 for Structure-Level Knowledge Distillation For Multilingual Sequence Labeling

* Accepted to ACL 2020, camera-ready. 14 pages

Access Paper or Ask Questions