I am a 5th year PhD student at the Language Technologies Institute in the School of Computer Science at Carnegie Mellon University. My advisor is Prof. William Cohen.

My primary research interest is in the use of machine learning for network analysis and text mining. Specifically, I am interested in latent topic and blockstructure analysis using topic models and stochastic block models. My research is motivated by a need to address the disquietude brought about by the overwhelming complexity of reality. Graphical models allows us to explain and understand complex real world phenomena using grok-able approximations.

About me

I am a news junkie who grew up in Bangalore, India where the dosas are reason enough for living. After graduating from PESIT with a Bachelor's degree in Computer Science and Engineering in 2003, I worked on text categorization and information extraction problems at the Applied Research division at Yahoo! for four years before starting in the PhD program at the LTI. My very supportive wife and I have been living in Pittsburgh since 2007; ergo I am a Steelers fan in addition to being a fan of the Indian cricket team.

In my non-researcher avatar, I try to practice Vipassana meditation and adopt a minimalist lifestyle.

Research Projects

Most recently, I have been developing techniques to regularize mixed membership stochastic models to incorporate some of the advantages of spectral clustering methods into stochastic network models.

We have been colloborating with political scientists to model decision making in politics by analyzing political blogs.

In 2008 and 2009, I worked on include developing nonparametric topic models for text with temporal variation.

In 2007 and 2008, I worked on the Message Task Linking (aka MeTaL) part of the RADAR project during which Vitor Carvalho and I developed CutOnce - a Thunderbird extension to do recipient recommendation and leak detection.

Recent Activities

Teaching Assistant for Language & Statistics I - Spring 2011

Teaching Assistant for Machine Learning 10-601 - Spring 2010

Intern at Google Pittsburgh during the summer of 2009 where I worked on Product Search.

Webmaster for ICWSM - 2009 and 2010.

Publications

2011

What pushes their buttons? Predicting comment polarity from the content of political blog posts [pdf | bib]
Ramnath Balasubramanyan, William W. Cohen, Doug Pierce and David Redlawsk.
In Workshop on Language in Social Media (LSM 2011) at ACL-2011

Combining stochastic block models and topic models [pdf | bib]
Ramnath Balasubramanyan and William W. Cohen.
In SDM 2011, SIAM Conference on Data Mining.

A shorter version was presented at the ICML 2010: Workshop on Topic Modeling [pdf | bib].

2010

Node Clustering in Graphs: An Empirical Study [pdf | bib]
Ramnath Balasubramanyan, Frank Lin and William W. Cohen.
In NIPS 2010: Workshop on Networks Across Disciplines in Theory and Applications

From Tweets to Polls: Linking Text Sentiment to Public Opinion Time Series [pdf | bib].
Brendan O'Connor, Ramnath Balasubramanyan, Bryan R. Routledge, and Noah A. Smith.
In ICWSM 2010: Proceeedings of the fourth International AAAI Conference on Weblogs and Social Media.
Press coverage - Pittsburgh Tribune-Review, CNN Tech, Economic Times BBC Radio 5 (at 13:00) and others.

2009

From Episodes to Sagas: Understanding the News by Identifying Temporally Related Story Sequences [pdf | bib]
Ramnath Balasubramanyan, Frank Lin, William Cohen, Matthew Hurst and Noah A. Smith
In ICWSM '09: Proceeedings of the third International AAAI Conference on Weblogs and Social Media (Poster).

Information Leaks and Suggestions: A Case Study using Mozilla Thunderbird [pdf | bib]
Vitor Carvalho, William Cohen and Ramnath Balasubramanyan
In CEAS '09: Conference on Email and Anti-Spam

2008

Activity-centred Search in Email [pdf | bib]
Einat Minkov, Ramnath Balasubramanya and William Cohen
In CEAS '08: Conference on Email and Anti-Spam

CutOnce- Recipient Recommendation and Leak Detection in Action [pdf | bib]
Ramnath Balasubramanyan, Vitor Carvalho and William Cohen
In The AAAI 2008 Workshop on Enhanced Messaging

Earlier

Document preprocessing for naive Bayes classification and clustering with mixture of multinomials [pdf | bib]
Dmitry Pavlov, Ramnath Balasubramanyan, Byron Dom, Shyam Kapur and Jignashu Parikh
In KDD '04: The tenth ACM SIGKDD international conference on Knowledge discovery and data mining (Poster).