Titre : | Deep models for understanding and generating textual arabic data |
Auteurs : | Mohamed Berrimi, Auteur ; Mohamed Saidi, Directeur de thèse |
Type de document : | document électronique |
Editeur : | Sétif : Universite ferhat abbas faculté des sciences de l’ingénieur département d’informatique, 2023 |
ISBN/ISSN/EAN : | E-TH/2254 |
Format : | 1vol.(171 f.) / ill.en coul |
Note générale : | Bibliogr. |
Langues: | Anglais |
Catégories : | |
Mots-clés: | Textual arabic data ; Multiple challenges of arabic |
Résumé : |
In this thesis, our primary focus lies in advancing the research efforts and making significant contributions to the Arabic language domain through the utilization of the state-of-the-art (SOTA) deep learning techniques. Another specific objective is to enhance Arabic resources by creating diverse Arabic text and speech corpora encompassing various dialects that are relevant to a variety of NLP tasks. In addition, we delve into the exploration and development of effective models specifically designed for Arabic language processing applied for the aforementioned NLP tasks. These models were tailored to perform consistently in both Modern Standard Arabic (MSA) and dialectal Arabic datasets. By developing and evaluating these models and their applications, this thesis contributes significantly to the field of Arabic NLP, paving the way for future research and advancements in solving the unique challenges of processing Arabic text and speech. |
Côte titre : | E-TH/2254 |
En ligne : | http://dspace.univ-setif.dz:8888/jspui/bitstream/123456789/4261/1/Final-Thesis-Mohamed-Berrimi.pdf |
Exemplaires (1)
Cote | Support | Localisation | Disponibilité |
---|---|---|---|
E-TH/2254 | Thèse | Bibliothèque centrale | Disponible |
Accueil