I mostly work on language learning and machine learning. My recent paper is on second language acquisition modeling while using the Duolingo Application. This paper was part of the shared task called SLAM 2018 organized by Duolingo at the 13th Workshop on Innovative Use of NLP for Building Educational Applications.

The shared task was to predict the likelihood of a user making an error or a mistake while learning a new language. My approach in SLAM 2018 relied on insights from research in code-switched text for language learning. I hypothesized that, while answering, users make use of contextual clues such as the sound and part-of-speech of the surrounding words. My approach finished 9th in the competition.

Here is the link -

http://www.aclweb.org/anthology/W18-0524