The following repo provides a deep NER model to identify POS from subtitles text samples.
(1) data.processing.r - This script will process the gz file of subtitles which can be downloaded from http://opus.nlpl.eu/OpenSubtitles.php and provide a csv data file for further modeling stored in the csv folder.
(2) dner.py - This script will train a deep NER model on the finaldf.csv, attempting to accurately idenfity POS from subtitles.