Deep Learning Methods for Social Media Text Mining by Jeremy Bertolet '24

Deep Learning Methods for Social Media Text Mining by Jeremy Bertolet ’24, Wednesday April 24, 1:10 – 1:50pm, North Science Building 015, Wachenheim, Statistics Colloquium

Abstract: Social media websites such as X (formerly Twitter), Instagram, Reddit, etc. are utilized globally, and the mining of social media texts can provide valuable insights on consumer behavior, sentiment analysis, and natural disaster resource allocation among other things. Two commonly used deep learning methods in social media text mining are Convolutional neural networks (CNN) and Long short-term memory (LSTM) networks. In this talk, we will go into more detail on the mechanics of these two types of networks and briefly dive into their applications in a case study for natural disaster resource allocation.