Social Computing with Yuheng Hu at EASM 2015
Social Computing with Yuheng Hu at EASM 2015 Heading link
Professor Yuheng Hu recently returned from the 2015 IEEE International Conference on Data Mining, where he organized and chaired EASM 2015, the first-ever international workshop on social media-based event analytics. For Hu, whose work lies at the intersection of social science, machine learning and data analytics, the event gave a big boost to the emerging field of social computing.
“The sheer magnitude of content generated by social media outstrips traditional social science approaches – this is where social computing comes in,” says Professor Hu. Hu applies big data techniques to questions about human behavior on social media. In which contexts do people share their location? In which contexts do people interact more with others who are geographically nearby? Questions like these underlie much of Hu’s research.
“One area of interest for me relates to first-screen, second-screen behavior. Basically, what do people post about what they see on tv?” In his research, Hu used the example of Twitter activity during the Romney-Obama debates.
“The easy assumption is, tweets will mirror the broadcast across time. What we observed, however, were before and after periods when tweets referred to the debates very generally, and then a period during the debate itself when tweets became specific, referring to healthcare, jobs, foreign policy, and other issues raised by the candidates, though not in the same order as the debate. During the debate, discussion of these issues on Twitter jumped around at random.”
In his teaching, Hu shows students pathways into the seemingly limitless trove of data available for research. “We live in the era of big data. Students need hands-on experience.” Hu starts students with the fundamentals of data management and analytics, and moves on to machine learning and specific data-driven questions.
“I’m a news junkie, and social media is where you see responses to the news. So even the simple question of how people respond to a given news story – this involves analyzing a huge swath of behavior as it varies over time. The analysis puts all the methodology into action.”
Asked about his recent move to UIC, Hu replied, “I love the diversity of the student body, which brings so many perspectives into the classroom. And the support of fellow faculty is great. But I have to mention, after so many years in Arizona and California, I’m finally seeing all four seasons!”