Algorithms based on subregular hypothesis for induction of phonological grammars and sets of underlying forms from morphophonological paradigms. (In progress.)
ostia.py, fst_object.py, and helper.py courtesy of @alenaks's SigmaPie package.
The Simplex Input Strictly 2-Local Decomposition Learning Algorithm of Hua & Jardine 2021 is implemented in si2dla.py.
The Input (Strictly Local) Decomposition Learning Algorithm (To Appear) is implemented in idla.py.
Other files are experimental variations on this algorithm:
| File | Description |
|---|---|
so2dla.py |
An Output Strictly 2-Local version of the SI2DLA |
fsi2dla.py |
A featural version of the SI2DLA |
features.py |
Some code to work with features |
kcxt.py |
Classify PBase pattern data by k-contexts |
The file testing.py contains some test data sets; this can be run from the command line to see how the algorithms perform.
This work is supported by NSF Grant #2416184.