PhD Researcher | Artificial Intelligence & Natural Language Processing
Connecting language research with applied AI and real-world data challenges.
Fedwa Mansouri is a doctoral researcher in Computer Science at the University of Gabès and a member of BECOME Institute's scientific team through the becomePhD programme. Her work connects artificial intelligence, natural language processing and machine learning with practical data-driven applications.
Her doctoral research explores the development of language models using Deep Learning and Reinforcement Learning. It focuses on optimising model parameters and training data as part of an ongoing investigation into language-model capabilities.
She holds a research master's degree and a bachelor's degree in Computer Science from the Institut Supérieur d'Informatique et de Multimédia de Gabès, Tunisia.
Her experience as a Data Scientist includes preparing and cleaning textual data and developing NLP models to classify banking transactions from their descriptions. This work connects Python development and machine learning with the practical requirements of data quality and automatic categorisation.
Her earlier research examined intrusion detection in vehicular networks using Federated Learning and Blockchain, alongside questions of cybersecurity, privacy and homomorphic encryption.
Through becomePhD, Fedwa brings a focus on language technologies, scientific experimentation and software development to BECOME's applied research community.
Her participation reflects the institute's founding commitment to supporting emerging talent, sharing knowledge and connecting research with practical challenges across borders. It also contributes to BECOME's research direction in Responsible AI & Scientific Computing, where mathematical methods, data and application knowledge are brought together.
Her published research explores federated learning and security, including intrusion detection in vehicular networks and privacy-preserving approaches to learning from medical data. Explore the publication records below for their original authorship, research context and publisher references.
2 records
This study examines a distributed intrusion-detection approach for vehicular ad hoc networks using federated learning and blockchain. It connects collaborative model training with security and privacy questions in connected transport systems.
View detailsThis conference paper examines federated learning with homomorphic encryption for medical-data privacy. It addresses how collaborative model training can be studied alongside the protection of sensitive data.
View detailsSupporting selected doctoral research through sponsorship, professional mentoring and academic collaboration, with a commitment to shared knowledge and useful applications.
View details