L Venkata Subramaniam 

Senior Technical Staff Member and Senior Manager - AI Science
Member IBM Academy of Technology
IBM Master Inventor
Location Head, IBM Research India, Delhi

IBM Research - India

4, Block - C, ISID Campus
Vasant Kunj, New Delhi - 110 070, INDIA

Phone: +91 11 4129 2201
FAX: +91 11 2613 8889
email: lvsubram_AT_in.ibm.com 

I was born in New Delhi, India. I received the B.E. degree in Electronics and Communication Engineering from Mysore University, the M.S. degree in Electrical Engineering from Washington University, St. Louis, USA, and the Ph. D. degree in Electronics from the Indian Institute of Technology, New Delhi, India. I joined IBM Research India in 1998 where I currently lead a team of world class researchers across India, Singapore and US developing the next generation AI technologies. I am an IBM Master Inventor.

I enjoy teaching and interacting with students. I often teach at IIT Delhi and was on the senate at IIIT Delhi from 2014-2016. I also like writing on technology in everyday life. My monthly series Folk Technology appeared in the i.t. magazine in 2008.

Research Interests

  • AI, Neuro-Symbolic Reasoning, Noisy Data Analytics, Unstructured Information Management, Statistical Natural Language Processing, Text and Data Mining, Information Theory, Speech Technologies.

AI Reasoning:

I work to enable machines to use and ingest background knowledge from disparate domains to drive comprehensive reading, reasoning, and question answering. The long-term goal is to build tools and techniques that minimize the manual effort required in both knowledge acquisition and reasoning to accomplish tasks in disparate domains like healthcare, finance, IT support. I am developing a general-purpose pipeline for knowledge acquisition and reasoning that provides the ability to acquire knowledge from disparate knowledge sources and apply it to novel domain-specific tasks enabling superior solutions involving learning of complex concepts and tasks.

Noisy Text Analytics:

Text produced in informal settings (email, blogs, tweets, SMS, chat) and text which results from processing (speech recognition, OCR) is inherently noisy. Operations like IR, IE, Mining, NLP need new techniques to contend with the noise. Over the years, apart from text, I have searched for informative patterns in signals ranging from speech, image and text to communication pulses. My team's work on Noisy Text Analytics has been recognized as an IBM research accomplishment in 2011. Read more about my research here.

FAQ Retrieval Task at FIRE 2013:

SMS Based FAQ Retrieval Task, Forum for Information Retrieval Evaluation (FIRE), New Delhi, Dec 4-6, 2013

Tutorial on Reliability Aware Data Fusion:

Reliability Aware Data Fusion, COMAD, Pune, Dec 15, 2012

AND 2012 Workshop along with COLING:

Workshop on Analytics for Noisy Unstructured Text Data (AND), Mumbai, India, Dec 9, 2012

Tutorial on Noisy Text Analytics:

Noisy Text Analytics, NAACL HLT, Los Angeles, 1 June 2010 [pdf presentation slides]

IJDAR Special Issues on Noisy Text Analytics:

Vol 14, No 2, June 2011: Guest Editors: L. V. Subramaniam, D. Lopresti, S. Roy, K. Schulz

Vol 12, No 3, September 2009: Guest Editors: D. Lopresti, S. Roy, K. Schulz, L. V. Subramaniam

Vol 10, No 3-4, December 2007: Guest Editors: C. Knoblock, D. Lopresti, S. Roy, L. V. Subramaniam

Read my Counter Terrorism Series:

Fighting terror series


* IBM Outstanding Technical Achievement Awards 2011, 2012, 2015, 2017

* Research Division Award 2014, 2019

* IBM Master Inventor 2013 -

* IBM Seventh Patent Plateau 2013

* IBM Research Accomplishments 2009, 2010, 2011, 2013, 2014, 2019

* IBM Bravo Awards 2005, 2007