Every semester, successful businesses and institutions – big and small – turn to Cornell Information Science master’s students to help conceptualize new ideas, design solutions for inefficiencies, and deliver new products.
This collaboration between companies and students is the essential, culminating component of Info Sci’s Master of Professional Studies (MPS) program. Called the MPS Project Practicum (INFO 5901), the course is the ultimate, hands-on learning experience, as student teams apply knowledge and skills in research, design, front- and back-end development and more to help real companies solve real problems.
In the Spring 2019 semester, Huge Inc. – a leading creative agency based in Brooklyn – sought to develop a visualization tool to help client companies build stronger brand awareness and improve customer experiences. Online customer reviews can provide companies critical insight into what works and what doesn’t. The challenge, though, comes in synthesizing those reviews across the vast, online landscape into a single, eye-pleasing dashboard, all in a few mouse clicks.
Inside of a single semester, four Info Sci MPS students did just that.
Combining natural language processing and deep-learning architectures, the MPS team built a fully automated machine-learning system to analyze massive volumes of publicly available reviews and produce key visualizations for Huge and its client brands.
Huge team members were Info Sci students Jeffrey Tsang, James Zhan, Dylan Limnavong, and Amarachi Emeziem. Greg Lennartz, a lead data scientist at Huge, served as team advisor.
2019 Spring MPS Project Team Huge.JPG
Tsang tackled the neural network optimization piece, the text classifiers, and used state-of-the-art tensor flow models; Limnavong built a summarization component that automatically parceled through text and analyzed quotes; Emeziem tested classifications, and Zhan built the entire dashboard.
“Start to finish, a process that normally would take weeks of discovery, research and user testing, we reduced to a couple of hours,” he said. “At the end of the process, we get a fully automated and built PDF that we can send to internal teams or to a client. The team did absolutely amazing work. I’m super proud of them, particularly with a very, very difficult subject matter – from mathematics, to coding and all the different languages. They crushed it.”
In total for the Spring 2019 semester, MPS students completed projects for nine clients, including Red Hat and Workday.
Elsewhere, thanks to another MPS project team, a small but hugely popular deli in St. Louis now has the foundation for a mobile ordering app.
As team member Song Ye explained, small, local restaurants often can’t afford – or refuse outright – to pay the fees associated with mobile food-ordering platforms like Grubhub and DoorDash, let alone pay for their own customer app, like major brands can.
For Gioia’s Deli, a century-old St. Louis deli famous for salami sandwiches, Ye’s team designed the framework for a customizable, mobile-ordering app to help the deli compete with popular third-party delivery apps. The hope, he said, would be to offer the same low-cost, mobile framework for participating small, local restaurants.
“We’re trying to empower small business to have their own app platform,” he said. “Right now, we’re cooperating with Gioia’s Deli, but maybe in the future, we can adopt the same system for different small, local restaurants so they can have their own app systems with customer rewards, online ordering and even delivery.”
The Gioia’s Deli project team members were Kaitlyn Son, Carrie Huang, Jenny Lei, Daphne Sun, Rashaad Ahmad, and Song Ye.
The MPS Project Practicum is a semester-long, required project for students in Cornell Information Science’s Master of Professional Studies program. In the past, clients have included Google, Microsoft Research, Verizon, Workday, Assurant, and Credit Suisse. If you represent a company interested in proposing an MPS project, please email the MPS project instructor at is-mps-projects@cornell.edu.