MIT EECS Undergraduate Research and Innovation Scholar
An Improved ConceptNet5 Build System
The first part of my research will be to determine the best solution and put into production an improved build system for ConceptNet5. The current process requires manual intervention and can take more than 24 hours. The new build system should be as automated as possible and significantly quicker. Once this system is in place, I will work to improve ConceptNet5s generation of natural language. This will make human interaction with ConceptNet5 easier, and allow for it to extract more information from less structured text sources.
Worked at Qualcomm Labs to help monetize new developing technologies. Worked at TripAdvisor developing classification algorithms to help verify that users only see valid content. Worked at CVISION Technologies, where I improved the companies website by adding non-english content for consumers from non-english markets.