Yahoo! Research Academic Visitor Program
Back in mid-2005, when Yahoo! Research was getting off the ground, Prabhakar Raghavan and Ron Brachman had a brainstorm. Why not establish interactions with respected colleagues in academia as a way to create collaborations with faculty, establish a potential pipeline for new hires and interns, and communicate to the academic community what Yahoo! Research was all about.
Since formalizing the Academic Visitors Program, Yahoo! Research has hosted dozens of leading academics from many of the top universities across the globe. These well-regarded thinkers have come to Yahoo! Research for anywhere from one week to one full year to share their insights, collaborate with scientists, and put their ideas into action.
"It’s very natural for an industrial research group like ours to reach out and interact with the academic community," Brachman says. "We are doing deep scientific research, and the closest analogs to our people tend to be university faculty."
Academic colleagues are always welcome to drop by Yahoo! Research to give a talk or spend the afternoon, but the Academic Visitors Program works a little differently. At the beginning of each year, Yahoo! research scientists submit their nominations for potential visitors, including the value that visitor could add to Yahoo!.
These nominees are then reviewed, prioritized and ultimately selected by an executive committee based on several informal factors, including depth of expertise, potential for substantive collaboration, and area of focus, with added weight given to those fields where Yahoo! Research does not have in-house experts but needs to get up to speed.
By engaging with the academic community, Yahoo! researchers are better able to vet their own ideas through peer review, and gain a clearer understanding of how their work fits into a broader context. It also allows Yahoo! to learn from the latest and greatest thinking on university campuses.
"The goal is to multiply our brain power and extend our intellectual community," Brachman says. "We want to make sure there are more people out there who are thinking about the kinds of problems that we at Yahoo! find very important."
This thought is echoed by Yahoo! Research scientist John Langford, who recently spent two weeks hosting Alex Smola from National ICT Australia. "The program helps us inform the rest of the world about what are interesting problems," Langford says. "This helps both Yahoo! and the academic visitor. Many academics are extremely intelligent, but are not aware of real-world problems. They can help solve these problems, and the solutions can be applied to Yahoo!."
For example, Smola and Langford focused their thinking on exploration problems, where only incomplete knowledge of the loss of choice is available. "Many of Yahoo!’s problems have this form and it is almost never studied in academia," Langford says.
Preston McAfee, vice president at Yahoo! Research, is currently gearing up to host David Reiley, a professor of economics at the University of Arizona who is coming for the year. Reiley started the branch of economics known as field experiments, where randomized trials are carried out in the real world. For example, he sold trading cards online through different auction formats to assess which format worked best in practice.
"I am very excited to have David come to Yahoo!," McAfee says. "He has been extraordinarily entrepreneurial and I hope to see him invent experiments that let us learn more about our customers. From his perspective, we have lots of properties that can be improved by experiment."
Another advantage of the program is that it can be used as an effective recruiting tool. "We can leverage the personal connection with academics for recruiting and investing in some long-term and more open-ended projects," says Yahoo! Research scientist Ravi Kumar, who recently hosted David Liben-Nowell from Carleton College. "Plus, it helps us identify and bring in good summer students."
As for the academic visitors, they love coming to Yahoo! because they can work closely on real-world applications in ways they can’t always do on their own campuses. These leading-edge thinkers typically live in a world where they believe their ideas should apply to a company like Yahoo!, but they can’t really test their theories.
By working closely with Yahoo! Research under the right agreements, they receive access to data, systems and technical talent they could not get anywhere else. Indeed, Yahoo! has such an extraordinary amount of widely varying data, it’s almost like hitting the jackpot for these faculty members.
The Academic Visitors Program is continually exposing Yahoo! Research to some of the most advanced and important ideas out there. "We cannot be close-minded about other work going on," Brachman concludes. "The real risk with a well-funded research lab like ours is to believe your work is the best no matter what else is happening outside. This program keeps us open to new ideas and allows us to obtain honest appraisals of our own work."
