I haven't posted about my research in a while (because pandemic, whee), but I'm pleased to share "Graph-theoretic Properties of the Class of Phonological Neighbourhood Networks", to be presented at CMCL: aclweb.org/anthology/2021…
The paper outlines the beginning of a research agenda in the formal properties of phonological neighbourhood networks, which is a representational tool for looking at how lexical structure is organized.
Studies involving phonological neighbourhood networks are on the rise, yet we (as a field) don't know very much about the intrinsic properties of these networks. This could lead to problems, especially if we just blindly apply standard network-theoretic methods.
So, this paper is an attempt to begin to solve the problem, using graph theory. I'm not a mathematician and can only go so far, but I think there's a decent stab at the problem here. Still lots more to do, however.
(And yes, there's an egregious typo at the bottom of the second page. "E(H)⊆V(G)" should read "E(H)⊆E(G)".)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
