I study
computer science
at
Columbia University's
School of Engineering and Applied Science.
This is a picture of my face.
I'm interested in the relationship between computation and
natural language, in both formal and applied settings. My
current research involves statistical machine translation and
language use classification.
I'm president of the
CU Association for Computing
Machinery chapter and and mentor Bronx high-schoolers with
2Train Robotics.
I used to do
mock trial.
I've worked for
the Lawrence Berkeley Laboratory,
Microsoft Live Labs
and Google.
Contact:
- jda2129@columbia.edu (COMS-3261- or ESP-related)
- jdandreas@gmail.com (everything else)
- My public key
I like to make things. Computers help. (Check out my github!) In no particular order:
Web
- CUGenerals (mentioned on TechCrunch)
- canvascity, a location-aware collaborative paint application.
Mobile
Desktop
- Firefox geolocation for Linux (still better than the one that ships with FF now, although it doesn't work with recent FF builds)
- The core Gnome Do Microblogging plugin
AI (NLP, ML, vision)
- Slides from my HackNY Fall '11 presentation (predicting the popularity of an article based on its title).
- Student, a JavaScript machine learning library [winner of the (tongue-in-cheek) "Best Javascript Machine Learning Library Award" at HackNY Spring '11 and demoed at NYTM]
- Tomato as input device (interestingly, Microsoft came out with a very similar demo only a few days after I put this on YouTube)
- OpenCV webcam X pointing device
- Experiments in augmented reality
- PHILOMENA, a declarative language for automatic counterpoint composition
Physics-ish
- The Atlas Multimedia Educational Lab for Interactive Analysis
- The Particle Data Group's publication management system
Graphics & UI
- The Columbia University Mock Trial website.
- A litestep theme from high school, with features a really cool dock that I still want to rewrite for a Linux DE some day.
People I have worked with: Ron Gejman, Hans Hyttinen, Ben Mann. You may also be interested in my research.
Courses
- Sp12: Computer Science Theory (COMS 3261)
- Sp11, Fa11: Emerging Scholars Program (COMS 3998)
Workshops
Tutoring
I occasionally offer private tutoring services. Please contact me via email for more information.
High School
- I'm a 2Train Robotics programming coach (FIRST 395). We won the 2011 York Regional! Watch my students' software autonomously hang an inflatable tube.
I work with the Columbia Natural Language Processing Group (occasionally at the Columbia Center for Computational Learning Systems). I'm writing my senior thesis on learning a lambda-calculus interlingua for machine translation. Past research has included a parse fuzzification algorithm to improve syntactic reordering for Arabic-English SMT, and a classifier for identifying agreement and disagreement in threaded discussions.
Publications
- Fuzzy Syntactic Reordering for PSMT. Jacob Andreas, Nizar Habash and Owen Rambow. In proceedings of the EMNLP Workshop on Machine Translation, July 29, 2011, Edinburgh, Scotland. paper, slides
- Corpus Creation for New Genres: A Crowdsourced Approach to PP Attachment. Mukund Jha, Jacob Andreas, Kapil Thadani, Sara Rosenthal and Kathleen McKeown. In proceedings of the NAACL Workshop on Creating Speech and Language Data With Mechanical Turk, June 6, 2010, Los Angeles, California. paper
- Semi-Automated Annotation for Prepositional Phrase Attachment. Sara Rosenthal, William J. Lipovsky, Kathleen McKeown, Kapil Thadani and Jacob Andreas. In proceedings of the 7th International Conference on Language Resources and Computation (LREC), May 2010, Valletta, Malta. paper
Coursework
- On the hardness of learning in the "Principles and Parameters" framework. For Intro. to Computational Learning Theory (COMS 4252) with Rocco Servedio. paper
People I have worked with at Columbia: Mike Collins, Sara Rosenthal, Kathy McKeown, Nizar Habash, Owen Rambow. My Erdös number is at most five. You may also be interested in my personal projects.