Exploiting natural neural control to minimize speech errors among language learners

点击翻译按钮获取中文摘要

Learning a new language is hard, especially for adults. While vocabulary, grammar, writing, reading andlistening are all challenging, learning to speak in a new language is hardest of all. This is particularly true whenlearning a language very different than one's own.This project would circumvent various difficulties inherent in learning to speak in a foreign language. We willconstruct a software platform able to autonomously guide language learners towards proficiency in a foreigntongue, either independently or with the support of a human tutor. The project would implement discoveriesabout how humans learn new languages, along with powerful algorithms that can analyze and manipulate eachlearner's attempts to speak in a foreign language.The project is a collaboration between Dr. Frank Rudzicz (Assistant Professor, Computer Science, Universityof Toronto) and Dr. Simon Overduin (Founder, Syngli Inc.). Dr. Rudzicz is a leading research within computerscience, and has made important contributions to the understanding of speech analysis, recognition, featureextraction and manipulation. Dr. Overduin's own research background focused on the optimization and neuralunderpinnings of behaviour, including speech. His start-up Syngli applies some of the research by Dr. Overduinand colleagues, in the form of an intelligent tutoring system that is able to track and guide learning in a widevariety of domains.