CLARIN-PL
Not a member yet
504 research outputs found
Sort by
Speech Recognition System for Polish: Studio Quality
No full textThis resource contains dockerized models and scripts of an automatic speech recognition system for Polish trained on studio quality speech. The system is based on the Kaldi toolkit. The scripts include methods for performing speech recognition, forced alignment and a lenient alignment of audio. The Github repository contains information on how to use the tool
PELCRA EMO corpus
No full textThe corpus comprises 30 focused structured interviews (17 hours and ca. 200000 word tokens) centred on the topic of emotions. The corpus has bibliographic, morphosyntactic and sociolinguistic annotation
Word Sense Disambiguation tool WoSeDon
No full textWord sense disambiguation tool using rich, semantic resources
KPWr annotation guidelines - normalisation of temporal expressions
No full textKPWr annotation guidelines - local and global normalisation of temporal expression
PELCRA PARL corpus
No full textThe corpus comprises 50 sampled recordings (12 hours) and manual transcriptions (ca. 101 00 word tokens) of parliamentary data
Speech Recognition System for Polish: Parliamentary Speech
No full textThis resource contains dockerized models and scripts of an automatic speech recognition system for Polish trained on Polish Parliament speeches. The system is based on the Kaldi toolkit. The scripts include methods for performing speech recognition, forced alignment and a lenient alignment of audio. The Github repository contains information on how to use the tool
Bulhakov - corpus of events, temporal expressions and temporal relations
No full textThe corpus contains the text of the short story "Fatalne jaja" by Michaił Bułhakov (http:// www.wolnelektury.pl). The corpus is manually annotated with temporal expressions, events and temporal relations using TimeML gudelines adapted to Polish