Institute of Information Science, Academia Sinica

With the maturity of location-acquisition techniques and mobile devises, a novel type of online services, geo-social media (e.g. Foursquare, Facebook, and Flickr), have catched not only users' attention but also researchers' efforts on solving real-world problems. People are allowed to keep track of what they see, where they go, who they know, what they experience, and when they do anything in the forms of images, locations, social network, short texts, and time stamps respectively. Heterogeneous data in geo-social media can be considered as a kind of "sensors" that reflect how human beings interact with the physical environment in the world. In this paper, we seek to point out and discuss a new research direction that connects geo-social media sensor data to enable novel environmental applications. Specifically, we target at introducing three environmental tasks and preliminarily solving these tasks by exploiting textual, geographical, social, and temporal data in geo-social media . For the first and second parts, while air pollution and noise pollution in urban areas are more and more severe and damages the public health, we study how air quality can be accurately inferred and predicted at any locations and at any time using geo-social media data. By analyzing city's road net, traffic flows from government agencies, and geographical venues and short texts in geo-social data, we find the these features play a significant role on reflecting real-world air quality index and noise degree. The third part is to reveal how heterogeneous geo-social media data can be applied for business environments. Using check-in records in Twitter, and place information and online social relationships in Foursquare, we point out that it is possible to predict the commercial foot traffic anywhere and anytime in urban areas, and find both visual and geographical factors are the most deterministic. In order to deal with these three tasks, we develop a general-purpose environmental information inference framework to not only model user-environment interactions, but also is preliminarily validated to be capable of accurately infer air quality, noise degree, and commercial foot traffic. Such promising results are believed to encourage a brave new direction for the media communities to devote their future efforts and enable more real-world environmental applications from geo-social media.

Hsun-Ping Hsieh is now an Assistant Professor at Electrical Engineering Department at National Cheng Kung University. He received his Ph.D. degree in Graduate Institute of Networking and Multimedia at National Taiwan University, and his M.S. degree in Information Management at National Taiwan University as well. His research interests include big data mining, urban computing, geo-social computing, and geographical information systems. His international recognition includes ACM KDD Cup 2010 First Prize (member of NTU team),Garmin Fellowship 2014 and 2013, Best Intern Award at Microsoft Research Asia in 2013 and NTU Outstanding College Youth in 2014. In the past years, he published a series of papers in location-based services and urban computing, including TIST, WWW, KDD, PKDD, SocialCom, ICWSM and CIKM. He had been the main tutorial presenter on the topic of social media analytics at WWW 2015, and on route planning at ICWSM 2014 and ASONAM 2014.