Updated README.md for 2023

This commit is contained in:
Aarne Ranta
2023-03-21 07:46:59 +01:00
committed by GitHub
parent 1332d550cd
commit 7e9e636776

View File

@@ -5,13 +5,6 @@ Therefore we will structure it according to the exercise sessions
rather than chapters.
The abstract syntax is given in the subdirectory grammars/abstract/
The compulsory lab assignment has two alternatives:
1. Complete MiniLang for your language of choice
2. Complete MicroLang + application (Doctor) for your language of choice
If you want to work on the Mini grammar you can ignore everything about the Micro grammar and just work on the corresponding Mini file instead
## Session 5
1. Design a morphology for the main lexical types (N, A, V) with parameters and a couple of paradigms.
@@ -28,40 +21,16 @@ If you want to work on the Mini grammar you can ignore everything about the Micr
## Session 7
1. Complete the rest of the definitions of RGL categories and functions in MiniGrammar.
2. Add concrete UD labels.
3. Generate a synthetic UD treebank. The simplest way to do this from the
GF shell is random generation; change english/MiniLangEng.labels to
the labels of your own language:
1. Add concrete UD labels.
2. Generate a synthetic UD treebank by using gfud.
gr -cat=Utt -number=22 -depth=6 | vd -output=conll
-abslabels=abstract/MiniLang.labels -cnclabels=english/MiniLangEng.labels
*Hint*: if your language is already in the standard RGL, you can implement it easily by making a copy of the files in grammar/swedish.
But use this implementation only as a reference with which to test your own one: your lab submission must not make any use of the RGL modules, except Prelude!
*To deliver* (but only if you select this assignment): file Mini..X.gf for your language X, together with MiniLangX.labels and a treebank in CoNLL format with 20 trees.
*To deliver* for your language X, file MicroLangX.labels and a treebank in CoNLL format with 20 trees.
*Deadline*: until the end of the course.
## Session 8
### Alternative 1: Doctor
1. Implement a simple application grammar, application/Doctor.gf.
This can be be done either by using your own MiniResource or the standard resource.
There are two versions of English implementation to support this.
Your first approximation can be just to copy it and change the words.
*To deliver* (but only if you select this assignment): file DoctorX.gf for your language X.
*Deadline*: until the end of the course.
### Alternative 2: Abstract Wikipedia
The task is to add a language to the grammar in `wikipedia/`. This involves
- finding country and other names to implement `CountryNames`, by querying Wikidata labels and using `extract_names.py`;
- writing the concrete syntax modules for `Facts`, `CountryNames`, and `Countries` for this language.
The task is to add a language to the grammar in `wikipedia/`. Mode details coming soon.
More information on the task can be find in
- the lecture notes: https://github.com/aarneranta/NLG-examples/blob/main/doc/gf-nlg.pdf
@@ -69,7 +38,7 @@ More information on the task can be find in
- the GF Summer School film: https://www.youtube.com/watch?v=gX_y2BqJ0w0&list=PL7VkoRLnYYP6EZngakW7lNNCTjfC93uh0&index=15
## A method for testing your Micro and Mini grammar
## A method for testing your Micro grammar
Since MicroLang is a proper part of the RGL, it can be easily implemented as an application grammar.
How to do this is shown in `grammar/functor/`, where the implementation consists of two files: