Current progress
- A quick update on the inductive bias in lexical tone project. I reviewed all publicly available corpora, and only the Tong corpus explicitly annotates lexical tone production errors. I’m now testing alternative aligners to see if I can automatically extract potential tone errors from the other corpora. So far, I’ve identified 175 disyllabic tone errors across two corpora.
Meeting notes from Aug 29th
- 2 papers I need to look at
- CL-child corpus: https://ieeexplore.ieee.org/document/10800018
- corpus of lexical tone errors: https://hal.science/hal-03435803v1/document
- possible relook at the aligner