Running CLAN POS tagger for human validation

exporting .eaf files as .cha files with CLAN

1. have .eaf files in a directory
2. set "working" on command window for clan to the directory
3. from clan, run elan2chat @ to create .cha files for all .eaf files in the same directory

adding %mor tier with CLAN

1. have .cha files in a directory
2. set “working” on command window for clan to the directory
3. from clan, file —> get mor grammar —> English
4. run “mor *.cha” to run MOR, POST, POSMORTEM, and MEGRASP in sequence —> adds %mor and %gra to files

clean_cha_file.py

correcting for cut-offs with - not getting recognized

• USE: python3 clean_cha_file.py cha cleaned_cha
• NOTE: have cha files in eafs_cha directory first
• OUTPUT: iterates through cha files and removes word-final dashes; saves output in cleaned_cha folder

clan_mor.py

getting cha file to a readable form with tags

• INPUT: takes a .cha file as input
  • have to add %mor tier first
• NOTES: finds the %mor lines and corresponding speaker; output is in chronological order
• OUTPUT: file_name line_number speaker utterance token tag readable_tag correct_tag corrected_tag
• sample output lines:

COU_DI46_030001_10800000_11100000_random-20.cha 18 CHI I want to xxx . � want v verb COU_DI46_030001_10800000_11100000_random-20.cha 21 MA1 it's not in there Bud . � there n noun