Discovering the City by Mining Diverse and Multimodal Data Streams
February 24th Tuesday, 2015 16:00-17:00
Room 1208, National Institute of Informatics Access
Professor Winston H. Hsu Graduate Institute of Networking and Multimedia & Department of Computer Science and Information Engineering, National Taiwan University http://www.csie.ntu.edu.tw/~winston
Due to the explosion of social media, people are used to sharing theirdaily life on the Internet. Over the decade, there have been many socialservices developed for different user groups. Based on thecharacteristics within, people will choose different media to sharetheir thoughts, photos, personal preferences, etc. For ex- ample, peoplemay utilize Foursquare for recording places where they go, Instagram forsharing their visual experiences via photos or videos, and Twitter forupdating things that interest them. Therefore, these social mediacapture different moments in the daily life.
Instead of relying on single source where most prior works arerestricted to, we aim to leverage multiple data streams including socialmedia and open data for mining urban activities. We argue a systematicframework, which detects rich and multi-angle semantics and attributesto facilitate further applications. Moreover, we devise advanced methodsto leverage complementary information across sources and deal with thenoises in such user-contributed data streams.
In this talk, we will highlight on: 1) leveraging the complementaryinformation for data analytics, 2) discovering the latent behaviors,synergies and differences among media sources, 3) mining effectivefactors parameterizing emerging applications from heterogeneous andmulti-granularity signals, 4) devising detection and learning methodsrobust to noisy user-contributed data, 5) measuring user engagements inInstagram by more than 40 million photos across eight internationalcities, and 6) identifying the opportunities and challenges for miningdiverse data streams.
Dr. Winston Hsu is an active researcher dedicated to novel algorithmsand systems for ultra-large-scale image/video retrieval, social mediamining/recommendation, multimedia analytics, and mobile and cloud-basedmultimedia applications. He is keen to realizing advanced researchestowards business deliverables via academia-industry collaborations. Heis an Associate Professor in the Department of Computer Science andInformation Engineering, National Taiwan University, also a VisitingResearcher at Microsoft Research Redmond (Summer 2014). He received hisPh.D. degree (2007) from Columbia University, New York. Before that, heworked for years as a founding engineer in the multimedia softwarecompany, CyberLink Corp., where experiencing Engineer, Project Leader,and R&D Manager.
He has been recognized with technical awards and best paper awards inmultimedia research community (e.g., ACM Multimedia 2006 Best PaperRunner-Up, First Prize in ACM Multimedia Grand Challenge 2011, FirstPlace in MSR-Bing Image Retrieval Challenge 2013, ACM Multimedia2013/2014 Grand Challenge Multimodal Award, etc.). He delivered keynotespeeches and lectured several highly rated and very attended tutorialsin top conferences (e.g., ACM Multimedia 2008/2009, SIGIR 2008, andICASSP 2009/2011). He is on the Editorial Board for IEEE MultimediaMagazine and IEEE MSA TC. He is in the organizing committee for ACMMultimedia (2010/2012/2013/2016), ICMR 2015/2016, ICME 2009, PCM2013/2014, MMM 2013. He is in the technical committee for majorconferences (ACM Multimedia 2009-2015, SIGIR 2011/2013/2015, IJCAI 2015,WWW 2011, CIKM 2011, ICME 2006-2015). He is a Senior Member of both theIEEE and the ACM. He was awarded 2011 Ta-You Wu Memorial Award, anational and prestigious recognition for young researchers, and 2013National Outstanding IT Elite Award, for his contributions in advancedresearch and industry collaborations.