By Arnav Choudhury, Biomedical Engineering Student – Senior, UC San Diego. Most of us have found ourselves speaking to tens of people at a time and wondering, “Am I going too fast or too slow? Are they getting me?” Now imagine if you were speaking to 100 or more people for about an hour. And to make it worse, you know that four out of ten members of the audience are disengaged at any point in time. This is exactly what every professor teaching a large undergraduate class faces, three times a week.
Students too! When I first came to UC San Diego as a freshman from India, I was scared. I used to be in high school classes with fifty students, and I thought that was big. But at UCSD I was only one of 300. It was hard to ask questions or let the professor know I didn’t understand something.
This is why some other students and I built Mesh, a real-time feedback platform that helps professors improve their presentation and teaching so that they can ensure their messages get conveyed to students. We want to make professors’ lives easy while also making large lectures more interactive and engaging!
Designed as a plug and play solution for any device, it provides professors with real-time and recorded feedback about student understanding in class, indicating the parts of the lecture where students are most confused. During class, the platform sends professors notifications about student comprehension (via the emoji above) so they can provide immediate clarification.
It also recommends questions to ask students to check understanding, and enables professors to quiz students or take opinion polls. Using advanced machine learning techniques, it picks out the four most representative answers from over hundreds of short-text student answers, enabling the professor to quickly uncover misconceptions.
Using the professor’s laptop microphone, it also records a digital copy of every lecture, combined with information about student comprehension, so that after class the professor may review and improve course content and delivery.
Mesh provides students with a way to get personalized attention in today’s overpopulated university classrooms. Using built-in machine learning algorithms, Mesh answers routine student questions automatically through a built-in chat bot. If they have additional questions, students can interact with the bot, as well as their teaching staff, via the app.
Over time, the bot can learn to answer more sophisticated student questions related to class content, reducing work for both professors and teaching assistants. Moreover, Mesh recommends questions for professors to ask students to further check their understanding, and lets professors quiz students with short text or multiple choice questions, or poll them for their opinions.
To get Mesh rolling, we participated in a hackathon in Spring 2016 at UCSD called Startup UCSD. We won an office space at the Qualcomm Institute Innovation Space on campus, which gave us a morale boost, and we went on to test Mesh with over 1500 students at UC San Diego.
Faculty have also participated in testing. Dr. Makeba Jones, assistant professor in the Education Studies Department, tested Mesh in the spring of 2016. She said, “Using Mesh was a great way to increase participation in my large university classes. It was more successful than encouraging oral discussion. Usually six to seven students participate in whole-class discussions regularly, whereas over eighty students participated in class discussion through Mesh. It was very successful!”
We are planning a third beta test this summer to further refine the platform, which will consist of even more artificial intelligence to suggest questions that students could ask professors in real-time, after all sometimes students don’t even know what they don’t understand, while also recommending out-of-class resources to students to improve their learning.
We’re excited to take Mesh further. With Mesh, professors become super professors!