Distributional semantics for lexical variation and change

• Event: Seminar
• Lecturer: Bill Noble
• Date: 27 January 2021
• Duration: 2 hours
• Venue: Gothenburg

Word representations based on the distributional hypothesis are useful in a wide range of natural language prediction tasks. They have also been used extensively to study long-term lexical semantic change. In this talk, I will discuss the use of distributional semantics to study more short-term semantic change, as well as variation between speech communities. I will present results from two recent studies and compare the merits of using word vectors from a diachronic Skip-Gram model versus conditional neural language models.