Shashi NarayanResearch Associate
Informatics Forum 338
School of Informatics
The University of Edinburgh
10 Crichton Street
Edinburgh EH8 9AB, UK
Phone: +44 (0)131 650 4418
My research focuses on natural language understanding and generation. I aim to develop general frameworks for generation from underlying meaning representations or for text rewriting such as summarization, text simplification, question answering and paraphrase generation. The aim is to develop a general framework that is not specific to an application and which incorporates task-specific syntactic and semantic information. I mostly rely on deep learning and spectral methods to develop such framework. I also have experience with parsing and other structured prediction problems.
Previously, I was a doctoral student at Université de Lorraine, Nancy, working at Loria INRIA under the supervision of Claire Gardent. I was awarded an Erasmus Mundus Masters scholarship in Language and Communication Technology (EM-LCT). I did my major (Bachelor of Technology, Honors) in Computer Science and Engineering from Indian Institute of Technology (IIT), Kharagpur India.
- June, 2018
- I presented a poster for our paper Ranking Sentences for Extractive Summarization with Reinforcement Learning at NAACL 2018, New Orleans. Thanks for stopping by! The data, code and demo are available.
- We gave a tutorial on Deep Learning Approaches to Text Production at NAACL 2018, New Orleans. It was the second most registered tutorial at the conference. Thanks for attending! The abstract and slides are now available!
- May, 2018
- Congratulations to Joana! Her long paper "Local String Transduction as Sequence Labeling, Joana Ribeiro, Shashi Narayan, Shay B. Cohen and Xavier Carreras will appear at COLING 2018.
- April, 2018
- Our long paper "Document Modeling with External Attention for Sentence Extraction" with Ronald Cardenas, Nikos Papasarantopoulos, Shay B. Cohen, Mirella Lapata, Jiangsheng Yu and Yi Chang will appear at ACL 2018.
- The data and code for our NAACL 2018 summarization paper are now available at Github.
- We have upgraded the Split and Rephrase Benchmark (EMNLP 2017) to Version 1.0. It is now available at Github.