My new work uses subculture identity theory and looks for identity markers in discourse, both in news and in social media. The goal is to investigate how we construct who we are by stating who we are not.
I'm a social psychologist who studies belonging, mainly through discourse analysis. It has been fascinating to watch the emergence of a social science community here.
Next week UCU will serve notice on every university in the UK for 3 days of strike action in November. You can read more about the campaign here: https://www.ucu.org.uk/article/12469/FAQs
Great question on control of data projects when they've been released. Lauren mentioned a concept of "data retirement" which is hugely important and needs to be a standard norm in data analytics/data science.
In my data analytics class tonight we're discussing ML and text analysis. I'm going to show them how to classify record reviews with some diagnostics. The LDA classifier works pretty well. Rock reviews are the hardest to classify.