Adith Swaminathan

Graduate Student
Department of Computer Science
Cornell University

E-Mail: [first name]
Puzzle time in the Himalayas
4161 Upson Hall,
Ithaca, NY 14853-7501
Curriculum Vitae (updated December 2012)
Past projects
Pre-print drafts of conference papers

About Me

I am a second year graduate student in the Department of Computer Science at Cornell University. I am easily distracted by (cover-letter-speak: my current interests are) new or elegant ideas in Machine Learning (more broadly, Artificial Intelligence) and Game Theory (more narrowly, Multi-Agent Systems) with applications that grapple with ambiguity (Natural Language Processing) and/or adapt to an ever-changing worldview.
CS@Cornell has amazing faculty who are also sociable mentors; So far, I have interacted primarily with (non-cover-letter-speak: I ambush them with silly questions) Prof. Thorsten Joachims, Prof. Johannes Gehrke, Prof. Charles Van Loan and Prof. Éva Tardos. My graduate minor is in Applied Mathematics.
I interned with the Search Labs at Microsoft Research, Silicon Valley during the summer of '12, and worked as a strategist with Tower Research Capital for 14 months from 2010-11. I received a Bachelor of Technology degree in Computer Science and Engineering from Indian Institute of Technology, Bombay in 2010, interned over the summer of 2009 at Microsoft India Development Centre as a Software Development Engineer-Tester. Before that, I played games and sports all day every day.


Temporal corpus summarization using submodular word coverage. CIKM 2012
Ruben Sipos, Adith Swaminathan, Pannaga Shivaswamy and Thorsten Joachims.
Evolving corpora can be summarized through time-lines indicating the influential elements. We created interesting timelines of articles in the ACL anthology and NIPS conference proceedings over a span of 15 years using landmark documents, influential authors and topic key-phrases. We achieved this by formulating a sub-modular objective extending the coverage problem over words (traditional summarization) to incorporate coverage across time.

Beyond myopic inference in Big Data pipelines. KDD 2013
Karthik Raman, Adith Swaminathan, Johannes Gehrke and Thorsten Joachims.
Pipelines constructed using modular components have very brittle performance: errors in earlier components cascade through the pipeline causing catastrophic failures in the eventual output. We explored simple ideas from Probabilistic Graphical Models to make the inference procedure more robust, while still using off-the-shelf components to compose pipelines.

Scratchpad (mostly Vogon poetry)