Conclusion
- CAs are a theoretical model of neural processing where a CA represents
a concept.
- CAs can be thought of as computational primitives.
- These primitives can be used to represent concepts, hierarchies,
rules and structures.
- Processes are more complex, but the natural language parsing process
can be modelled using CAs.
- The parsing model is based on CAs that implement a stack to bind
relevant grammar rules and sentential constituents.
- CAs are learned implicitly.
- This CA-based parsing model is a nonconscious model of language
processing.
- CA modelling is currently relatively primitive. I am currently
exploring hierarchies and rules in a CA-based model.
