Sense.ly applet makes ratings personal
Friends’ opinions and relatives’ suggestions are often important factors in people’s everyday decisions. What if one application were able to synthesize all of their recommendations into one source through Facebook?
That is exactly what Jonathan Chu, a senior computer science major, and Anlu Wang, a junior computer science major, accomplished with their new social recommendation engine Sense.ly.
Todd Eisenberger, a senior computer science major, and Ryan Pearl, a junior computer science major, also aided in the project. A program called Project Olympus aided in Sense.ly's transition from the university to the business environment.
Chu originated the idea for Sense.ly in his everyday search for high-quality restaurants in Pittsburgh, especially since he often relied on the advice of his friends. “I figured I’d make it easy to ask friends for recommendations and then get a machine learning algorithm to verify them for you. I didn’t want to stop at restaurants, however, so we’ve added other categories such as movies, TV shows, etc., to help flesh out the types of recommendations,” Chu said via e-mail.
By visiting the website, students can sign up and use the app on Facebook. Then, recommendations will begin to show up in the student’s News Feed, based upon both friends’ ratings and the individual ratings that the student gives to restaurants and movies. Any ratings given by those in a person’s networks will appear as well.
“Whenever you rate items, it powers our machine learning algorithm. Telling Sense.ly about what you like and dislike lets us learn your tastes. The more things you rate, the more accurate our algorithm becomes,” Chu said.
Carnegie Mellon provided Chu and Wang with experience in algorithms and building large, well-designed systems. Chu especially recommended 15-251, Great Theoretical Ideas in Computer Science; 15-451, Algorithm Design and Analysis; and 15-410, Operating System Design and Implementation. “The computer science undergraduate program at Carnegie Mellon lays a solid and rigorous foundation in abstract, analytical, and creative thinking skills, transcending any technology fad of the day,” Jeannette Wing, head of the computer science department, said in an e-mail.
Project Olympus, founded in January 2007 by computer science professor Lenore Blum, seeks to provide support from the earliest stages of business development for both faculty and students at Carnegie Mellon. Members of the Sense.ly project took advantage of this opportunity.
“Project Olympus is a Carnegie Mellon innovation center that assists students and faculty in exploring the commercial potential of their research and innovations by providing mentoring, micro grants, startup advice, incubator space and connections for faculty and students across campus and with the wider regional, national and global business communities,” Blum stated. “A key goal is to augment and accelerate the process of moving groundbreaking research and great ideas to development and business stages. [Its] bustling incubator space for students is located off Craig Street.”
Developed projects are named PROBES, PRoblem-Oriented Business Explorations. Project Olympus offers guidance through its in-house business adviser, Kit Needham, and embedded entrepreneur, Babs Carryer.
To date, there have been 72 PROBES from across the university, of which 44 have formed into companies. Of the 44 companies, students formed 31 and faculty 13. At least 180 students have found guidance from the program since its origination. One of the first faculty PROBES to find business success was reCAPTCHA, a project led by Luis von Ahn, an assistant professor of computer science and now a member of the executive board of Project Olympus.
Sense.ly is currently in the PROBES stage in the Project Olympus program. Chu and the other creators of Sense.ly are working closely with the program and its advisers to further the project.
“Students work at their own speed. Some teams want to meet with me regularly,” Needham said via e-mail. “Others call upon me when they need advice, a contact or help with a specific problem. We provide basic education on the Dos and Don’ts of forming business teams, starting a business on an F-1 student visa, [and] legal workshops (in cooperation with University of Pittsburgh’s Innovation Practice Institute). These events also help students from various schools and backgrounds to connect with each other to form diverse business teams.”