Kashif Shah (Ph.D)

Senior Research Scientist at Amazon AI
San Jose, CA, USA
More About me
News:
Website is no more maintained

[March 8, 2016] Long paper on Multitask Learning for MT QE got accepted at @NAACL2016

[Dec 21, 2015] Paper accepted in ICASSP 2016: Raymond W. M. Ng, Kashif Shah, Lucia Specia, Thomas Hain; Groupwise learning for ASR k-best list reranking in spoken language translation

[July 29, 2015] Paper accepted in EMNLP 2015: Shah, Kashif; R.W. NG, Bougares Fethi, Specia, Lucia; Investigating Continuous Space Language Models for Machine Translation Quality Estimation”

[Jun 10, 2015] Our paper on the robustness of ASR rerank for spoken language translation got accepted in Interspeech 2015

[Jan 31, 2015] Paper published in MT journal: Shah, Kashif; Cohn, Trevor; Specia, Lucia;" A Bayesian non-Linear Method for Feature Selection in Machine Translation Quality Estimation, MT Journal, 2015.

About Me

I am working as a Senior Research Scientist at Amazon AI exploring machine learning methods for NLP. I spent 2.5 years at eBay as Research Scientist and then as a Research Manager. Before that, I was a Postdoctoral research associate in Natural Language Processing group at Department of Computer Science, University of Sheffield working primarily with Prof. Lucia Specia. My research revolved around Statistical Methods to Natural Language Processing and Machine Learning. I have been involved in many EU machine translation projects (e.g. QT21 and Cracker). I did my PhD under the supervision of Prof. Holger Schwenk and co-supervision of Dr. Loïc Barrault at LIUM, University of Le Mans, France where I worked on domain / model adaptation techniques to improve machine translation systems. I have contributed in many open source implementation such as QuEst.

Research

Research Interest

In General, my research focuses on natural language processing and exploring / applying machine learning techniques to language technology especially machine translation. I have worked on various topics but more notably these are: Statistical Machine Translation, Quality estimation, Language and Translation modeling, Model Adaptation, Word Embedding and (Deep) Neural Networks.

Research Projects

I am named researcher in following ongoing projects:
2015- QT21 and Cracker (Both granted by EU under Horizon 2020)
Completed Projects:
2012-2014 QTLaunchPad (EU) , Okapi-QuEst (Sponsored by EAMT) and QuEst (PASCAL2 - Harvest program)
I have also participated in following projects during my PhD:
2009-2012 EuroMatrix/EuroMatrixPlus, DARPA GALE , COSMAT projects

Professional Activities

I have been reviewing for ACL, EMNLP, COLING, NLE JOURNAL, WMT, MT-SUMMIT, EAMT
I am member of ACL, EAMT and NLP & MT group, University of Sheffield.
I have participated in various shared Tasks: NIST OpenMT-09/12, WMT-11/12/13/14/15, IWSLT-09/11/14

Publications (Peer Reviewed)

The USFD SLT System for IWSLT 2014

Raymond W. M. Ng, Mortaza Doulaty, Rama Doddipatla, Wilker Aziz, Kashif Shah, Lucia Specia, Thomas Hain, Oscar Saz, Madina Hasan and Ghada Al-harbi; in IWSLT , 2014

Publications (Others)