CourseMatch: Using Word Embeddings for Scaling Access to Online Education during COVID
At Coursera our mission is to provide everyone, anywhere access to high quality education to transform their lives. As the COVID crisis forces teaching and learning to shift online, there is an acute need to support both universities and students at this time. To enable universities and students to take advantage of the large number of high quality, online courses on Coursera, we have utilized word embeddings to map on-campus university catalogs to their online Coursera equivalents, ensuring students can continue their studies and universities can support their programs. During this presentation we will review the theory behind world embeddings and discuss this novel application in greater detail.
Vinod Bakthavachalam is a Data Scientist working with the Content Strategy and Enterprise teams, focusing on using Coursera's data to understand what are the most valuable skills across roles, industries, and geographies. Prior to Coursera, he worked in quantitative finance and studied Economics, Statistics, and Molecular & Cellular Biology at UC Berkeley.