Merge pull request #12 from opensource-spraakherkenning-nl/add-variat…

…ions Add some variations + remove dash (-) again
opensource-spraakherkenning-nl · Nov 6, 2023 · 86ec141 · 86ec141
2 parents 0016129 + 221eeeb
commit 86ec141
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 1 deletion.
diff --git a/ASR_NL_benchmark/normalize.py b/ASR_NL_benchmark/normalize.py
@@ -44,7 +44,7 @@ def replace_numbers_and_symbols(text):
     >>> replace_numbers_and_symbols('12,3%')
     'twaalf komma drie procent'
     """
-    removed_punct = string.punctuation.replace("'", '').replace("-", '')
+    removed_punct = string.punctuation.replace("'", '')
     text_without_symbols = replace_symbols(text)
     clean_text = replace_numbers(text_without_symbols)
     clean_text = clean_text.translate(str.maketrans('', '', removed_punct))

diff --git a/ASR_NL_benchmark/variations.glm b/ASR_NL_benchmark/variations.glm
@@ -13,6 +13,8 @@ z'n => zijn / [ ] __ [ ]
 'k => ik / [ ] __ [ ]
 'r => er / [ ] __ [ ]
 'ns => eens / [ ] __ [ ]
+ie => hij / [ ] __ [ ]
+da's => dat is / [ ] __ [ ]
 d'ruit => eruit / [ ] __ [ ]
 restaurant- => restaurant / [ ] __ [ ]
 jeugd- => jeugd / [ ] __ [ ]