Analysis of PhilPapers

In my talk, I will present an analysis of a corpus of contemporary philosophical texts that I am working on with my colleagues Kaše and Švadlenková. The analyzed corpus is based on the PhilPapers database (https://philpapers.org) and consists of 67,653 English-language documents, the majority (70%) of which are academic articles. Our interest is the detection of metaphors related to artificial intelligence. To this end, we explored the use of the MelBERT language model (Choi et al. 2021) (https://arxiv.org/abs/2104.13615).