Avneesh Singh Saluja
Senior Research Scientist, Netflix

I am currently a Research Scientist at Netflix, where I work on problems in statistical natural language processing and machine learning.

(CV) (Google Scholar) (Github) (Linkedin)

Overview

I completed my Ph.D. from the Electrical and Computer Engineering (ECE) Department at Carnegie Mellon University (CMU), affiliated with the Language Technologies Institute (LTI) in the School of Computer Science, and Carnegie Mellon Silicon Valley (CMU-SV). My thesis was on low-dimensional context-dependent translation models in statistical machine translation, and my advisors were Professors Chris Dyer and Ian Lane. I have also worked closely with Joy (Ying) Zhang.

Publications

Journals

• Online Discriminative Learning for Machine Translation with Binary-valued Feedback. A. Saluja and Y. Zhang. In Machine Translation, Vol. 28 (2) pp. 69-90, 2014. (Springer subscription required unfortunately)

Conferences

• Simplify-then-Translate: Automatic Preprocessing for Black-Box Translation, S. Mehta, B. Azarnoush, B. Chen, A. Saluja, V. Misra, B. Bihani, and R. Kumar. In AAAI 2020, February 2020.
• Using Aspect Extraction Approaches to Generate Review Summaries and User Profiles, C. Mitcheltree, V. Wharton, and A. Saluja. In NAACL - Industry Track, June 2018.
• (Best Paper Runner-Up) Language Modeling with Power Low Rank Ensembles. A.P. Parikh, A. Saluja, C. Dyer, and E.P. Xing. In EMNLP, October 2014.
  • Supplemental
  • Code (C++)
• Latent Variable Synchronous CFGs for Hierarchical Translation. A. Saluja, C. Dyer, and S. Cohen. In EMNLP, October 2014.
  • Code (Python)
• Graph-based Semi-Supervised Learning of Translation Models from Monolingual Data. A. Saluja, H. Hassan, K. Toutanova, and C. Quirk. In ACL, June 2014.
  • Code (C++)
• Machine Translation with Binary Feedback: a Large-Margin Approach. A. Saluja, I. Lane, and Y. Zhang. In AMTA, October 2012.
  • Code (C++, branched from cdec)
• Age-Layered Expectation Maximization for Parameter Learning in Bayesian Networks. A. Saluja, P. Sundararajan, and O.J. Mengshoel. In AIStats, April 2012.
• Context-aware Language Modeling for Conversational Speech Translation. A. Saluja, I. Lane, and Y. Zhang. In Machine Translation Summit, September 2011.

Workshops

• Hierarchical Encoders for Modeling and Interpreting Screenplays. G. Bhat, A. Saluja, M. Dye, and J. Florjanczyk. 3rd Workshop on Narrative Understanding, June 2021.
• Graph-Based Unsupervised Learning of Word Similarities Using Heterogeneous Feature Types. A. Saluja and J.Navratil. In TextGraphs 8: Graph-based Methods for Natural Language Processing, October 2013.
• Semi-Supervised Learning with Cover Trees. A. Saluja and B. Kveton. In Big Learning: 2011 NIPS Workshop on Parallel and Large-Scale Machine Learning, December 2011.
• Automatic Identity Inference for Smart TVs. A. Saluja, F. Mokaya, M. Phielipp, and B. Kveton. In AAAI 2011 Workshop on Lifelong Learning, August 2011.

Preprints

• Paraphrase-Supervised Models of Compositionality. A. Saluja, C. Dyer, and J.D. Ruvini. arXiv:1801.10293, February 2018 (originally February 2015).
• Infinite Mixed Membership Matrix Factorization. A. Saluja, M. Pakdaman, D. Piao, and A.P. Parikh. arXiv:1401.3413, January 2014.

History

I received a B.Sc. (with Distinction) from the Electrical Engineering Department at Stanford University in 2007.

In a previous avatar, I worked as an Analyst for the Structured Equity Solutions (SES) Group at Goldman Sachs in both New York and Hong Kong (2007-2009). I was also a Product Manager at July Systems in Bangalore, India (2009-2010), where the product that I launched and managed won the "Best Mobile International" category at the 2010 Mobile Excellence Awards.

Before Netflix, I was a Research Scientist at Airbnb, where I was a founding member of the AI Lab.

---

Last updated: August 2021