Yi Luan Email: luanyi [at] google [dot] com Research Scientist Google AI Language News 05-Aug-19 Join Google! 10-Aug-18 Our paper Multi-Task Identification of Entities, Relations, and Coreference for Scientific Knowledge Graph Construction has been accepted by EMNLP 2018. 19-Feb-18 Attend SemEvel 2018 Task 7 (relation extraction for scientific papers) and get first and second place in Subtask A and B. 01-Sep-17 My intern work with MSR Multi-Task Learning for Speaker-Role Adaptation in Neural Conversation Models has been accepted by IJCNLP. 30-Jun-17 Our paper Scientific Information Extraction with Semi-supervised Neural Tagging has been accepted by EMNLP. 09-Feb-17 Attend SemEvel 2017 ScienceIE Task and get first place in Subtask A and B. 12-Sep-16 Finished a wonderful internship at Microsoft Research NLP Group on personalized conversational models.

About Me

I am a Research Scientist at Google AI Language. I did my Ph.D. in ECE department at University of Washington, advised by Prof. Mari Ostendorf and Prof. Hannaneh Hajishirzi. My research area is Natural Language Processing, mainly focus on information extraction, knowledge graph construction and text generation for conversations. Previously, I got my master degree in Hirose & Minematsu Lab at University of Tokyo on robust speech recognition. Here is my CV.

Publications

• Entity, Relation, and Event Extraction with Contextualized Span Representations
  David Wadden, Ulme Wennberg, Yi Luan, Hannaneh Hajishirzi [pdf][code]
  ACL, 2019

• PaperRobot: Incremental Draft Generation of Scientific Ideas
  Qingyun Wang, Lifu Huang, Zhiying Jiang, Kevin Knight, Heng Ji, Mohit Bansal, Yi Luan [pdf]
  ACL, 2019

• A General Framework for Information Extraction using Dynamic Span Graphs
  Yi Luan, Dave Wadden, Luheng He, Amy Shah, Mari Ostendorf, Hannaneh Hajishirzi [pdf][code]
  NAACL, 2019

• Text Generation from Knowledge Graphs
  Rik Koncel-Kedziorski, Dhanush Bekal, Yi Luan, Mirella Lapata and Hannaneh Hajishirzi [pdf][Rik's code]
  NAACL, 2019

• Multi-Task Identification of Entities, Relations, and Coreferencefor Scientific Knowledge Graph Construction
  Yi Luan, Mari Ostendorf, Hannaneh Hajishirzi [pdf][data and code]
  EMNLP, 2018

• The UWNLP system at SemEval-2018 Task 7: Neural Relation Extraction Model with Selectively Incorporated Concept Embeddings
  Yi Luan, Mari Ostendorf, Hannaneh Hajishirzi [pdf][code]
  Proceedings of SemEval, 2018

• Multi-Task Learning for Speaker-Role Adaptation in Neural Conversation Models
  Yi Luan, Chris Brockett, Bill Dolan, Michel Galley, Jianfeng Gao [pdf]
  IJCNLP, 2017

• Scientific Information Extraction with Semi-supervised Neural Tagging
  Yi Luan, Mari Ostendorf, Hannaneh Hajishirzi [pdf] [code]
  EMNLP, 2017

• Multiplicative Representations for Unsupervised Semantic Role Induction
  Yi Luan, Yangfeng Ji, Hannaneh Hajishirzi, Boyang Li [pdf]
  ACL, 2016

• LSTM based Conversation Models
  Yi Luan, Yangfeng Ji, and Mari Ostendorf [pdf][data]
  

• Efficient learning for spoken language understanding tasks with word embedding based pre-training
  Yi Luan, Shinji Watanabe, Bret Harsham [pdf]
  Interspeech, 2015

• Recognition of stance strength and polarity in spontaneous speech
  Gina-Anne Levow, Valerie Freeman, Alena Hrynkevich, Mari Ostendorf, Richard Wright, Julian Chan, Yi Luan, and Trang Tran [pdf]
  SLT, 2014

• Relating automatic vowel space estimates to talker intelligibility
  Yi Luan, Richard Wright, Mari Ostendorf, and Gina-Anne Levow [pdf]
  Interspeech, 2014

• Semi-supervised noise dictionary adaptation for exemplar-based noise robust speech recognition
  Yi Luan, Daisuke Saito, Nobuaki Minematsu, Keikichi Hirosei [pdf]
  ICASSP, 2014

• Performance improvement of automatic pronunciation assessment in a noisy classroom
  Yi Luan, Masayuki Suzuki, Yutaka Yamauchi, Nobuaki Minematsu, Shuhei Kato, Keikichi Hirose [pdf]
  SLT, 2012