Colloquium: Text analytics and data acumen

Nicholas Horton, PhD, Amherst College
ZSR Auditorium, Tuesday, January 23rd 11:00am

60 years ago, Fred Mosteller and David Wallace published a paper in JASA that made inferences in an important authorship problem: the Federalist papers. Notwithstanding this groundbreaking work, text continues to be an important example of unstructured data that has traditionally been nearly absent from the statistics curriculum in secondary schools and early post-secondary education, despite the importance that the written word has in all aspects of education and society. In this talk, Dr. Horton will review this work and legacy, describe a more modern classification problem, discuss how new learners approached text analytics, and conclude with some thoughts about statistics and data science amidst the advent of powerful generative AI models such as ChatGPT, GitHub Copilot (now available within Posit/RStudio) and Bard.

For the comfort of all, please do not wear scented products.