Event Registry dataset texts with OCR degradations and synthesised segmentation

This is the text of the Event Registrt dataset dedicated to fake news detection that has been updated to be used in event detection.

Rupnik, Jan, Andrej Muhic, Gregor Leban, Primoz Skraba, Blaz Fortuna, et Marko Grobelnik. 2016. « News Across Languages - Cross-Lingual Document Similarity and Event Tracking ». Journal of Artificial Intelligence Research 55 (janvier): 283‑316. https://doi.org/10.1613/jair.4780.

Miranda, Sebastião, Artūrs Znotiņš, Shay B. Cohen, et Guntis Barzdins. 2018. « Multilingual Clustering of Streaming News ». In 2018 Conference on Empirical Methods in Natural Language Processing, 4535‑44. Brussels, Belgium: Association for Computational Linguistics. https://www.aclweb.org/anthology/D18-1483/.

Guillaume Bernard. (2022). Event Registry titles only dataset with multiple extracted features (both sparse and dense) (1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.6630447

Some degradations are applied using the DocCreator [1] tool in order to degrade the text of the tweets and to reproduce some common errors found in OCRised documents [2].

[1]: Journet, Nicholas, Muriel Visani, Boris Mansencal, Kieu Van-Cuong, et Antoine Billy. 2017. « DocCreator: A New Software for Creating Synthetic Ground-Truthed Document Images ». Journal of Imaging 3 (4): 62. https://doi.org/10.3390/jimaging3040062.

[2]: Linhares Pontes, Elvys, Ahmed Hamdi, Nicolas Sidere, et Antoine Doucet. 2019. « Impact of OCR Quality on Named Entity Linking ». In Digital Libraries at the Crossroads of Digital Information for the Future, 11853:102‑15. Lecture Notes in Computer Science. Cham: Springer International Publishing. https://doi.org/10.1007/978-3-030-34058-2_11.

The results of the OCR degradations are as follow:

Event Registry CER/WER

Without
Character degradation
Phantom degradation
Bleed
Blur
All

Event Registry
CER
0.282
4.154
0.274
0.275
0.582
4.577

Event Registry
WER
0.552
16.364
0.551
0.548
1.159
16.974