We hope to make Dyna "strong enough for a researcher, sensitive enough for a student."
See our users page for a list of courses using Dyna.
Here's how you might use it in a class:
- Dyna makes it easy to teach the high-level structure of particular algorithms. For example, you can teach probabilistic parsing to linguistics students. (Faculty in at least two linguistics departments elsewhere have contacted us about doing this.) Note that Dyna supports visual browsing of the resulting proof forests.
- In a more computationally sophisticated class, the teacher is also free to delve into the details of how a Dyna program executes.
- In an NLP or AI class, Dyna provides a unified framework for implementation strategies that apply across multiple related or unrelated algorithms.
- Finally, students can use Dyna to tackle interesting problem sets or term projects on real data, in a reasonable amount of time.
You might also use Dyna without teaching it. For instance, it may be a good implementation language if you need to build some black-box systems for your students to experiment with.