Steven Wilson will give the talk, "Social Natural Language Processing for Personal Values, Human Activities, and Beyond."
Abstract: User-generated natural language content on the internet provides a massive opportunity for better understanding how users present what they are thinking, feeling, and doing in their everyday lives. Scientists interested in complex social, political, and cultural phenomena have turned to the content of online blogging and social media platforms as naturally occurring sources of observational data. In this talk, I will describe a collaborative effort between computer scientists and psychologists to measure expressions of personal values in text data and values’ connection to human activities, leading to new language resources and text analysis tools. After that, I will provide a few highlights from projects that leverage natural language processing methods within the domain of computational social science with topics ranging from emoji use to online slang. I will conclude with a call for future work with greater consideration of social context, social impact, and establishing user trust in these kinds of systems.
Steven R. Wilson is a Postdoctoral Research Associate and University Teacher (Instructor) in the School of Informatics at the University of Edinburgh studying applications of Natural Language Processing within Computational Social Science. He develops computational methods and tools for improved understanding and measurement of social/sociolinguistic phenomena and psychological constructs using data from sources including social media and free-response surveys. Previously, he has studied of personal values, human activities, and online slang, and currently is focusing on topics like misinformation, offensive language, humor, sarcasm, and emotion in online settings. His research has been published in ICWSM, ACL, WebSci, IJCNLP, and others, and his work on crowd-powered hierarchical lexicon development led to a best paper award at SocInfo in 2018. He currently teaches an upper-level course titled “Text Technologies for Data Science” at U of Edinburgh. Steve holds both a PhD and MS in Computer Science and Engineering from the University of Michigan where his dissertation was advised by Prof. Rada Mihalcea, and he earned a BS in Computer Science/Systems from Taylor University.
Meeting ID: 827 2371 0586
Questions? Contact Lori Kelso