TGLIS-227 is a sentence-level dataset for LIS. It is composed of
- trascript_news.csv contains the trascript of each the news
- topic_news.csv contains the topic of each news
- timestamp_audio_news.csv contains the timestamp of each audio news
- timestamp_video_news.csv contains the timestamp of each video news
By using timestamps it is possible to extract video and audio of each news from original newscasts.
Manuela Marchisio: [email protected], Università degli Studi di Torino - Corso Svizzera 185, 10149, Torino, Italy.
Alessandro Mazzei: [email protected], Università degli Studi di Torino - Corso Svizzera 185, 10149, Torino, Italy.
Dario Sammaruga: [email protected], Orbyta Tech S.r.l. - Piazza Castello 113, 10121 Torino, Italy.
All files and data is protected by Creative Commons Licence CC BY-NC-SA 4.0.
You're free to: use, distribute, communicate date in this repository in any format and transform it, under these constraints:
- Obligation to acknowledge proper attribution, provide a link to the license, and indicate if changes have been made;
- Prohibition of using the material for commercial purposes;
- Obligation to distribute contributions under the same license as the original material, even in cases where it has been transformed.
Link to licence terms: https://creativecommons.org/licenses/by-nc-sa/4.0/deed.en