Jonathan Huang: 2014 Networks of Shapes, Images, and Programs: The Power of Joint Data Analysis Workshop

 

Wednesday, April 16, 2014
Location: Fisher Conference Center, Arrillaga Alumni Center

"Data Driven Student Feedback for Programming Intensive MOOCs"
1:30pm - 2:00pm

Abstract:

In recent years an increasing number of students have turned to online resources, such as massive open online courses (MOOCs) for learning. But while these online courses give teachers more coverage, student-teacher ratios can often be ten thousand to one or worse. With such ratios, students no longer get the type of feedback they need to really understand the material.


Codewebs is a system that we have been developing which addresses the problem of scalability in providing student feedback for online programming-intensive courses. Codewebs analyzes a massive code corpora of historical student submissions and uses it to provide instant, useful and detailed student feedback to tens of thousands of students in the same course. By relying on a statistical approach, the quality of feedback increases as our system sees more data and the feedback is automatically tailored for each assignment. I will present a novel data driven technique to discover shared "parts" amongst multiple student submission, a problem that is complicated by the fact that there are always many ways to accomplish the same functionality in code. Throughout, I will demonstrate results on Coursera's Machine Learning course, which received over 1 million code submissions in its first run.


Bio:

Jonathan Huang is an NSF Computing Innovation (CI) postdoctoral fellow at the geometric computing group at Stanford University. He completed his Ph.D. in 2011 with the School of Computer Science at Carnegie Mellon University where he also received a Masters degree in 2008. He received his B.S. degree in Mathematics from Stanford University in 2005. His research interests lie primarily in statistical machine learning and reasoning with combinatorially structured data with applications such as analyzing real world education data. His research has resulted in a number of publications in premier machine learning conferences and journals, receiving a paper award in NIPS 2007 for his work on applying group theoretic Fourier analysis to probabilistic reasoning with permutations.