Data mining in the social-media ecosystem

RayÂmond Fu, a newly appointed assisÂtant proÂfessor of elecÂtrical and comÂputer engiÂneering, wants to build a better social-​​media ecosystem, one in which FaceÂbook makes expert friend recÂomÂmenÂdaÂtions and consumer-​​product suggestions.
"My goal is to bring the curÂrent social-​​networking system to the next level," said Fu, a machine-​​learning expert who will hold joint appointÂments in both the ColÂlege of EngiÂneering and the ColÂlege of ComÂputer and InforÂmaÂtion SciÂence. "Users," he added, "will defÂiÂnitely have access to a higher-​​quality serÂvice in the near future."
At the beginÂning of the year, Fu received funding from both a sciÂenÂtific research orgaÂniÂzaÂtion and a coaliÂtion of govÂernÂment agenÂcies to design algoÂrithms capable of rapidly anaÂlyzing conÂtent on social-​​networking and video-​​sharing webÂsites. In JanÂuary, for instance, he received a two-​​year, $360,000 postÂdocÂtoral research felÂlowÂship award from the U.S. IntelÂliÂgence ComÂmuÂnity, and in FebÂruary he received a three-​​year, $483,500 grant from the Air Force Office of SciÂenÂtific Research.
One algoÂrithm in progress, Fu explained, will be capable of preÂdicting the demoÂgraphics, behavÂioral tenÂdenÂcies and relaÂtionÂships between people who appear in photos and videos on FaceÂbook or YouTube. The other algoÂrithm will be capable of pinÂpointing their geoÂgraphic location.
"³§´Ç³¦¾±²¹±ô-​â¶Ä‹media users are sharing a lot of inforÂmaÂtion every single day," Fu explained. "Our goal is to develop an autoÂmatic method that could process this data very fast."
FaceÂbook adverÂtisers will have the ability to purÂchase the data for the purÂposes of tarÂgeting ads to parÂticÂular users. But, Fu explained, "Users would have to allow this conÂtent to be delivÂered to them." Issues of priÂvacy, he added, "would have to be resolved."
Both social-​​media projects doveÂtail with the university's focus on conÂducting use-​​inspired research that solves global chalÂlenges in cyberÂseÂcuÂrity. As Fu put it, "This research conÂsiders secuÂrity issues such as data comÂmuÂniÂcaÂtion, sharing and priÂvacy in cyberÂphysÂical networks."
Prior to joining the NorthÂeastern facÂulty, Fu served as an assisÂtant proÂfessor of comÂputer sciÂence and engiÂneering at State UniÂverÂsity of New York at BufÂfalo and was the founding director of its SynÂerÂgetic Media Learning Lab.
Fu hopes to set an ambiÂtious research agenda in his first year at NorthÂeastern. He looks forÂward to colÂlabÂoÂrating on interÂdisÂciÂpliÂnary research projects with sevÂeral facÂulty memÂbers, including Deniz Erdogmus, an elecÂtrical and comÂputer engiÂneering proÂfessor, and Marty Vona, an assisÂtant proÂfessor of comÂputer and inforÂmaÂtion science.
"NorthÂeastern has a very strong engiÂneering proÂgram," Fu said, adding that five gradÂuate stuÂdents and three postÂdocÂtoral researchers under his watch at SUNY BufÂfalo have transÂferred to the uniÂverÂsity. "There are many colÂleagues who share simÂilar research interÂests who I look forÂward to colÂlabÂoÂrating with."
Provided by Northeastern University