Dillon Dumesnil
MIT EECS — Mason Undergraduate Research and Innovation Scholar
Syntactical and Semantic Language Acquistion
2015–2016
Robert C. Berwick
The purpose of this SuperUROP is to expand on and improve the research on syntactical and semantic bootstrapping of how young children learn new verbs. The current framework of a problem is a sentence with a not real word as the verb (/He glipped the balloon/) and using Bayesian inference to determine which class of verbs glipped might be a part of. Giving a limited number of hypotheses the verb could belong to and through repeated use of the verb in a sentence, we are able to determine which hypothesis we believe is most likely. These inferences are then representative of how a young child (ages 3–10) will learn about new verbs. The goal is to raise the number of verbs we work with as well as be able to include a greater number of hypotheses to compare with the verb.
SuperUROP really interested me because I knew I wanted to do research while at MIT and this opportunity gives me the ability to spend an entire year on one project with the additional guidance of the 6.UAR lectures. I hope to learn more about what it is like to do research in Course 6 and I am excited the project also deals with learning and language development.