Retour aux résultats
Chapitres (2022), Natural Language Processing and Information Systems, Springer International Publishing, pp. 73-85

Uncertainty Detection in Historical Databases

Kouadri Wissam Mammar, Akoka Jacky, WATTIAU Isabelle , du Mouza Cedric

Historians analyze information from diverse and heterogeneous sources to verify hypotheses and/or to propose new ones. Central to any historical project is the concept of uncertainty, reflecting a lack of confidence. This may limit the scope of the hypotheses formulated. Uncertainty encompasses a variety of aspects including ambiguity, incompleteness, vagueness, randomness, and inconsistency. These aspects cannot be easily detected automatically in plain-text documents. The objective of this article is to propose a process for detecting uncertainty, combining dictionary-based approaches, and pattern identification. The process is validated through experiments conducted on a real historical data set. Lien vers l'article

KOUADRI, W.M., AKOKA, J., WATTIAU, I. and DU MOUZA, C. (2022). Uncertainty Detection in Historical Databases. In: Paolo Rosso, Valerio Basile, Raquel Martínez, Elisabeth Métais, Farid Meziane eds. Natural Language Processing and Information Systems. 1 ed. Springer International Publishing, pp. 73-85.

Mots clés : #Uncertainty, #Automatic-detection, #Historical-data, #Dictionary-Pattern