Actively Looking for Opportunities

Hello and welcome to my corner of the internet! I am Auday Berro, a computer scientist specializing in natural language processing (NLP) and Dialogue Services. My work is centered on the field of paraphrase generation and harnessing the potential of language models, including large language models (LLMs), to push the boundaries of what is possible in human-computer interactions.

Dr. Auday Berro received his PhD in Computer Science from Université Claude Bernard Lyon 1 (UCBL, France) on June 25, 2024, under the supervision of Prof. Boualem Benatallah. Since December 2024, he has been a postdoctoral researcher at the LIRIS laboratory in Lyon, France, under the supervision of Prof. Mohand-Saïd Hacid. His research focuses on AI-driven digital health, leveraging real-world patient data, artificial intelligence, and reinforcement learning to optimize sleep patterns and enhance mental well-being through adaptive, personalized interventions.
Before joining UCBL, he obtained a Bachelor's degree in Computer Science from the Lebanese University, Faculty of Sciences 4 (Zahlé, Lebanon) in 2017, followed by a Master's degree in Computer Science from the Lebanese University, Faculty of Sciences 1 (Hadath, Lebanon) in 2018. In 2019, he completed a Master 2 in Data Science for Risk Analysis, also at Faculty of Sciences 1, under the supervision of Prof. Ali Jaber. During this period, he undertook a research internship entitled Polluscope: microservices based data-intensive workflow, focusing on the automatic composition of Big Data applications using the cloud and containerization with Docker. This internship was carried out under the supervision of Dr. Yehia Taher within the ADAM team of the DAVID laboratory at the University of Versailles - Saint-Quentin-en-Yvelines (Versailles, France).

In the dynamic field of NLP, my research has focused on improving human-computer interaction through a better understanding of user requests in textual natural language. My PhD thesis, entitled "Paraphrase Generation for Conversational Services Learning", explores automated paraphrase generation techniques to improve training data for conversational AI, such as task-oriented chatbots. It presents ParapLine, a two-stage pipeline for generating diverse and relevant paraphrases by combining several generation methods and employing filtering mechanisms. This work also presents a taxonomy of errors identified in transformer-based paraphrase models and a manually annotated dataset for error analysis. In addition, it compares the effectiveness of large language models such as GPT-3.5 with crowdsourcing for the generation of syntactically diverse paraphrases, finding that LLMs offer a cost-effective alternative with superior syntactic novelty.

What I Do:

Key Contributions throughout my research:

My research interests include:

My journey into computer science began in early childhood, with a fascination for the development of video games. I was intrigued by the way games like Crash Bandicoot on the PlayStation 1 reacted to my joystick movements, and how the CD allowed the console to display the game on screen. This initial curiosity turned into a deep passion for the field. Lively discussions with my father and his friends, in particular Prof. Ir. Isin Argun, particularly about science and mathematics, fueled my interest. Over the years, I have worked on a variety of projects, from my undergraduate work at the Lebanese University to my doctoral thesis, each of which has strengthened my expertise and understanding of this ever-evolving field.

I believe in the power of knowledge sharing and collaboration. Whether you are interested in discussing potential projects, have questions about my research, or simply want to connect, feel free to reach out at audayberro@gmail.com.

Thank you for visiting my website. I look forward to connecting with you!