Who am i?
I'm Antoun J. Yaacoub, a Leading AI & Generative AI Expert.
Welcome to my digital space. As a Professor and Researcher, I specialize in Generative AI, Large Language Models (LLMs), and Data Platforms. With over 15 years of experience, I drive innovation and lead cutting-edge AI solutions from research to real-world application, while also nurturing the next generation of AI talent.
This site serves as a comprehensive overview of my professional journey, expertise, and contributions to the field of Artificial Intelligence.
If you have any questions, comments, or if you simply want to talk, please contact me.
Thank you for visiting!
Name: Antoun J. Yaacoub Ph.D.
Location: Paris, FRANCE
Email: [email protected]antoun.me
Key Facts Keys
Professional
- Scientific Coordinator, MSc AIBT at ESIEA
- Full-Time Associate Professor (AI/ML/Data Science) at ESIEA
- Applied AI Researcher & Innovator
- Leader in Generative AI/LLM Pedagogical Innovation
- Expert in End-to-End AI/ML Pipeline Development
- Academic Leader & Mentor in AI Disciplines
- Google Cloud Generative AI Leader
Personal
- Athlete
- Yogi
My Expertise Expertise
Areas of specialization - Pedagogic Innovation
Moodle Assessment Expert
Cambridge University / 2004 - 2007Highly experienced in educational design and tools, addressing distance learning through innovative teaching strategies.
- Designed and implemented innovative technical assessment tools (e.g., CodeRunner, AI-driven OneClickQuiz utilizing Generative AI and LLMs for personalized MCQ generation and question analysis), improving code quality, engineering efficiency, and adherence to applied AI research best practices.
- Developed advanced platform management strategies for technical curricula, including plagiarism detection and AI-driven content generation, to enhance instructional and operational effectiveness in AI-driven personalized learning development.
Moodle Trainer
Utilized Moodle for online education and training instructors to organize their course pages. Webinar presenter for the International Centre for Higher Education and Innovation (UNESCO-ICHEI) on topics such as AI-driven learning and evaluation.
Content creator and manager:
Managed IT systems, information platforms, and content creation (web pages, graphics, social media) for the Faculty of Science, enhancing communication and visibility. This included a focus on digital content strategy aligned with institutional goals.
Academic event planner, organizer and supervisor
Led agile planning, organization, and supervision of major academic and cultural events, fostering cross-campus collaboration and stakeholder engagement. Demonstrated strong leadership in team formation and task assignment to successfully complete large-scale institutional events.
Technical Expertise
Profound expertise in state-of-the-art AI/ML domains and technologies, including:
- AI/ML Domains: Generative AI (GenAI), Large Language Models (LLMs), Natural Language Processing (NLP), Information Retrieval (IR), Multimodal AI, Reinforcement Learning (RL), Supervised Learning (SL), Deep Learning, Transformers, Diffusion Models, Retrieval-Augmented Generation (RAG).
- AI/ML Paradigms/Concepts: AI Principles, Methodologies, Model Design, Evaluation Metrics, Fine-tuning Paradigms, SHAP, Model Tuning, Proof-of-Concept (PoC) Development, Research Translation.
- AI/ML Frameworks: PyTorch, TensorFlow, Scikit-learn, MLflow, LangChain, HuggingFace, numpy, keras.
- Cloud Platforms: AWS, Google Cloud.
- Programming Languages: Python (Proficient), R, SQL.
- Big Data & Data Technologies: Spark, Hadoop, PySpark, Large Datasets, Cloud-Native Environment, Data Structures, Algorithm Design.
- Methodologies: Agile (JIRA, Confluence), Product Lifecycle Management, Scientific Roadmapping, Experimentation, Causal Inference, Stakeholder Management, Communication.
Professional Experience
Scientific Coordinator, MSc AIBT
ESIEA / 2022 - PresentProvided visionary leadership and comprehensive technical management for the MSc AIBT program, guiding students in Generative AI and LLM applications for personalized learning transformation. Exercised budgetary oversight and resource optimization for program quality, demonstrating effective resource management for developing AI-driven personalized learning solutions. Designed and continuously updated academic programs focused on cutting-edge Generative AI, LLMs, and Data Platforms research, ensuring relevance and fostering experiential learning for building intelligent tutor systems and personalized learning experiences. Currently spearheading the strategic development and inaugural launch of a pioneering joint MSc program in AI, representing a bold international collaboration to expand global AI research in personalized learning and advanced model development for educational impact.
Full Time Associate Professor
ESIEA / 2022 - PresentDelivered a diverse range of AI/ML and Data Science courses, including Generative AI, LLMs, and Transformer Models, translating advanced research concepts into practical applications for students. Supervised numerous student AI/ML research projects, including hands-on model tuning, architecture design, and pipeline development, fostering practical problem-solving skills for building intelligent tutor systems and personalized learning applications. Pioneered and championed the integration of advanced techno-pedagogical methods and tools like CodeRunner and OneClickQuiz (leveraging Generative AI and LLMs for automated MCQ generation and in-depth question analysis), enhancing experiential learning in applied AI for educational content creation.
Google Cloud Certification Coordinator
ESIEA / 2023 - PresentFacilitated and coordinated cloud and Generative AI/LLM model platform certification programs in close collaboration with international partners, significantly enhancing technical skill development and supporting accessible learning experiences in AI technologies. Organized regular seminars on cutting-edge Generative AI/LLM/Cloud technologies, demonstrating responsiveness to innovation and fostering technical collaborations for upskilling teams in AI-driven educational product development.
Associate Professor
FACULTY OF SCIENCE, LEBANESE UNIVERSITY / 2012 - 2022Taught a wide range of computer science courses at the Applied Mathematics department, spanning from fundamental programming and data structures to advanced algorithms and competitive programming, focusing on student success and curriculum relevance for future careers in AI/ML.
Game Design and Development Master Supervisor
FACULTY OF SCIENCE, LEBANESE UNIVERSITY / 2017 - 2019Provided academic leadership and guidance for the Game Design and Development Master's program, overseeing curriculum design, content creation, and project management for industry relevance in AI/ML applications. Supervised, mentored, and provided academic guidance to master’s students, fostering their growth and career readiness. Established and expanded industry partnerships with local gaming companies to facilitate student internships, demonstrating effective network building and industry engagement in AI/ML fields.
Information Chief Officer
FACULTY OF SCIENCE, LEBANESE UNIVERSITY / 2014 - 2017Provided senior administrative oversight for IT infrastructure and systems, including technological guidance and vendor contract negotiations, optimizing resource allocation and operational efficiency for academic and research initiatives. Developed and implemented strategic and tactical goals for IT, ensuring IT alignment with broader institutional objectives and effective resource utilization. Managed staff and resources to effectively deliver and maintain technological solutions, supporting operational excellence and innovation in academic administration.
Cultural Service Chief Officer - Communication Officer
FACULTY OF SCIENCE, LEBANESE UNIVERSITY / 2017 - 2019Managed the Faculty's digital presence (website, social media), strategically enhancing dialogue with students and external stakeholders, including for program promotion and recruitment. Authored media releases, speeches, and newsletters, aligning communications with the Faculty's strategic vision and institutional policies. Led the strategic planning and execution of numerous high-profile on-campus and off-campus events (e.g., academic conferences, inaugurations), demonstrating strong network building and fostering community engagement. Ultimately, these comprehensive communication and event management initiatives were pivotal in strengthening the Faculty's external relationships, securing key partnerships, and driving significant institutional advancement and growth.
IT Manager
FACULTY OF SCIENCE, LEBANESE UNIVERSITY / 2011 - 2014Maintained information technology strategies by managing staff, researching and implementing technological strategic solutions.
Data analyst and Project manager
MIDDLE EAST STUDIES LTD / 2008 - 2014Collecting, collating and carrying out complex data analysis in support of management and customer requests, demonstrating analytical thinking for business decision-making.
My Certifications Certs
Google Cloud Generative AI Leader
Google Cloud / May 2025Demonstrated expertise in leading generative AI initiatives, including strategic planning, architecture design, and ethical deployment of large-scale generative AI solutions on Google Cloud. Validated proficiency in guiding teams through complex generative AI projects, leveraging Google Cloud's AI services and best practices to achieve business objectives.
Huawei Certification: Huawei Certified Academy Instructor
HUAWEI / 2020Authorized to deliver Huawei Authorized Information and Network Academy (HAINA) courses, demonstrating expertise in disseminating specialized technical knowledge.
Huawei Certification: HCIA - AI
HUAWEI / 2020Mastered the history of AI development, Huawei's full-stack AI system, and advanced algorithms in machine learning and deep learning. Proficient in building, training, and deploying neural networks using TensorFlow and MindSpore. Developed high-level competency in AI project management and technical support, demonstrating strong business acumen for AI applications.
Extracurricular Activity Extracurricular
My Teachings Teachings
Graphs & Data Concepts
Logic Programming & AI Concepts
Functional Programming
Web Development & Distributed Systems
Transversals & Research
Generative AI - ESIEA
AIBT MSC DEGREE / 2024-2025• Number of teaching hours: 18.
• Number of students: ~40.
Provided comprehensive instruction on the theoretical foundations and practical applications of Generative AI, encompassing models like GANs, VAEs, and Diffusion Models. Focused on experiential learning through projects involving image, text, and code generation, fostering innovation and a deep understanding of cutting-edge AI research.
Prompt Engineering - ESIEA
4th YEAR ENGINEERING DEGREE - AIBT MSC DEGREE / 2024-2025• Number of teaching hours: 6.
• Number of students: ~100.
Introduced students to the cutting-edge principles and advanced techniques of prompt engineering for Large Language Models (LLMs) and Generative AI. Emphasized experiential learning through hands-on prompt design, iterative refinement, and advanced strategies, preparing students for sophisticated AI interaction and future innovation.
Introduction to Computer Science - Lebanese University
BS 1ST YEAR COMPUTER SCIENCE - STATISTICS - MATHS 2015 - 2020• Number of teaching hours: 36.
• Number of students: 600.
Overview of computer architecture, operating systems, software, databases, networks, and the Internet, providing foundational knowledge for future computational studies.
Algorithms and Programming - Lebanese University
BS 1ST YEAR COMPUTER SCIENCE - STATISTICS - MATHS 2018 - 2021• Number of teaching hours: 60.
• Number of students: 300.
Covered fundamental elements of C programming, including basic types, operators, expressions, data input/output, alternative structures, and repetitive structures. Focused on practical application of algorithms and data structures.
Imperative Programming - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE - 2018-2019• Number of teaching hours: 40.
• Number of students: 120.
• Subbing Dr Siba Haidar in her sabbatical year.
Explored advanced imperative programming concepts including recursion, pointers, structures, linked lists, and file manipulation in C, strengthening students' ability to develop complex software solutions.
Data Structures - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE 2014 - 2021• Number of teaching hours: 60.
• Number of students: 150.
Taught fundamental concepts of data structures and algorithms, including asymptotic analysis, Abstract Data Types (ADTs), stacks, queues, trees, graphs, sorting, and searching, essential for efficient software development and data manipulation.
Advanced Algorithms and Competitive Programming - Lebanese University
BS 3RD YEAR COMPUTER SCIENCE - 2017 - 2021• Number of teaching hours: 40.
• Number of students: 20.
• Collaboration with Dr Siba Haidar
Covered advanced algorithmic techniques, including complexity of recursive functions, complete search, divide and conquer, greedy algorithms, dynamic programming, and advanced topics in graph theory, string processing, and computational geometry, preparing students for competitive programming challenges.
Programming with Python - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE 2020 - 2021• Number of teaching hours: 30.
• Number of students: 150.
Taught comprehensive Python programming, including control statements, functions, strings, lists, dictionaries, tuples, sets, files, regular expressions, lambda expressions, iterators, generators, and list comprehension, focusing on practical problem-solving in data science.
Numerical in Python - ESIEA
4th YEAR ENGINEERING - 2022-2025• Number of teaching hours: 18.
• Number of students: ~100.
Taught mathematical modeling and implementation of numerical methods in Python, including Gradient Descent, linear, and simplex methods, enabling students to formulate and solve optimization problems.
Image Processing - ESIEA
4th YEAR ENGINEERING - 2022-2025• Number of teaching hours: 12.
• Number of students: ~60.
Covered core image processing techniques including thresholds, histogram equalization, contrast enhancement, morphological transformations, filters, restoration, noise, segmentation, contours, descriptors, and face detection, with a focus on practical applications in computer vision.
Machine Learning - ESIEA
4th YEAR ENGINEERING - AIBT MSC DEGREE / 2022-2025• Number of teaching hours: 24.
• Number of students: ~100.
Taught principles of supervised learning, including algorithms like KNN, and neural networks (single and multi-layer, CNN), focusing on model optimization, performance measurement, and avoiding overfitting for robust AI inference.
Graph Theory and Applications - Lebanese University / ESIEA
BS 2ND YEAR COMPUTER SCIENCE - 2020 - 2021• Number of teaching hours: 30.
• Number of students: 120.
2ND YEAR ENGINEERING - 2021 - 2022
• Number of teaching hours: 48.
• Number of students: 28.
Provided deep knowledge of graph theory concepts and their applications in mathematics, natural science, and computer science. Covered graph representation, implementation, traversals, MST, coloring, Eulerian and Hamiltonian graphs, and shortest paths, enabling students to model complex problems.
Data Compression and Encoding - ESIEA
2ND YEAR ENGINEERING - 2022-2023• Number of teaching hours: 18.
• Number of students: ~33.
Taught fundamental concepts of information transport, main data encoding and compression techniques (lossless and lossy), enabling students to design and implement efficient compression algorithms adapted to various problems.
Introduction to Logic and Logic Programming - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE - 2013-2015• Number of teaching hours: 30.
• Number of students: 120.
Covered foundational logic (propositional and predicate) and Prolog programming, providing students with essential tools for formal reasoning and problem-solving in AI and computer science.
Advanced Databases - Lebanese University
MASTER 1 COMPUTER SCIENCE - 2013-2018• Number of teaching hours: 48.
• Number of students: 30.
Covered advanced database concepts: system crash recovery, logs, transactions, concurrency control, distributed databases, indexing (B-trees, B+trees, hashing), query optimization, deductive databases, and object-oriented databases (Oracle database management).
STRI PROFESSIONAL MASTER 1 - 2013 - 2015
• Number of teaching hours: 36.
• Number of students: 10.
Advanced topics in PL/SQL, including principles, exception handling, cursors, triggers, collections, records, dynamic SQL, Bulk SQL, procedures, functions, and packages, for robust database programming.
Knowledge Representation - Lebanese University
MASTER 1 COMPUTER SCIENCE - 2013 - 2017• Number of teaching hours: 36.
• Number of students: 20.
Taught fundamental concepts of knowledge representation, including predicates, conceptual graphs, ontologies, and web semantics, essential for building intelligent systems and semantic web applications.
Functional Programming - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE - 2018 - 2021• Number of teaching hours: 30.
• Number of students: 50.
Covered comprehensive Python functional programming, including control statements, functions, strings, lists, dictionaries, tuples, sets, files, regular expressions, lambda expressions, iterators, generators, and list comprehension, applied to data processing and problem-solving.
Graphical Interface and Applications - Lebanese University
BS 3RD YEAR COMPUTER SCIENCE - 2017-2018• Number of teaching hours: 36.
• Number of students: 50.
Taught design patterns (creational, structural, and behavioral) and JavaFX for building robust and scalable graphical user interfaces and applications.
Server Side Web Development - Lebanese University
BS 3RD YEAR COMPUTER SCIENCE - 2017-2018• Number of teaching hours: 50.
• Number of students: 50.
Covered web architecture, PHP, server-side scripting (forms, sessions), database connectivity (MySQL), Ajax, jQuery, security (injections), XML, XSLT, and regular expressions for building dynamic and secure web applications.
Distributed Applications Programming - Lebanese University
MASTER 1 COMPUTER SCIENCE - 2012 - 2019• Number of teaching hours: 48.
• Number of students: 40.
Taught principles of distributed applications, including Java streams, TCP sockets, UDP datagrams, multithreading, RMI, serialization, and agents, enabling students to develop robust networked systems.
Systems Administration in Networks - Lebanese University
STRI PROFESSIONAL MASTER 1 - 2013 - 2015• Number of teaching hours: 48.
• Number of students: 10.
Taught practical skills in network systems administration, including Windows server installation/configuration, server roles/features, virtual machines, DHCP/DNS services, Active Directory management (users, groups, OUs, GPOs), and firewall configuration.
Pattern Recognition - Lebanese University
SIGNALS, TELECOMS, IMAGE AND SPEECH (STIP), FIELD ELECTROMECHANICAL ENGINEERING (FEME) AND CYBERSECURITY (CBRS) MASTER 2 - 2021• Number of teaching hours: 8.
• Number of students: 36.
• Collaboration with Dr Siba Haidar and Dr Yasser Mohanna.
Taught practical applications of CNN for image classification and handwritten digit recognition, focusing on applied pattern recognition problems.
Scientific English - Lebanese University
GAME DESIGN AND DEVELOPMENT, CYBERSECURITY AND BIOSTATISTICS MASTER 2 - 2016-2017• Number of teaching hours: 36.
• Number of students: 30.
Taught scientific communication skills, including planning/delivering presentations at conferences, writing research reports, and creating professional CVs/LinkedIn profiles using LaTeX, essential for academic and professional dissemination.
Initiation to Scientific Research - ESIEA
5th YEAR ENGINEERING DEGREE - CYBERSECURITY MAJOR / 2022-2023• Number of teaching hours: 6.
• Number of students: ~90.
Introduced students to fundamental principles of scientific research, including bibliographic studies, academic writing (LaTeX), presentation skills (Beamer), and peer review processes, preparing them for academic and professional research contributions.
Introduction to Computer Science - Lebanese University
BS 1ST YEAR COMPUTER SCIENCE - STATISTICS - MATHS 2015 - till 2020• Number of teaching hours: 36.
• Number of students: 600.
• Content: Overview of computer architecture, operating systems, software, databases, networks and Internet.
Algorithms and Programming - Lebanese University
BS 1ST YEAR COMPUTER SCIENCE - STATISTICS - MATHS 2018 till 2021• Number of teaching hours: 60.
• Number of students: 300.
• Content: Fundamental elements of a C program. Basic types, operators and expressions in C. Read and write data. Alternative structure.
Imperative Programming - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE - 2018-2019• Number of teaching hours: 40.
• Number of students: 120.
• Subbing Dr Siba Haidar in her sabbatical year.
• Content: Recursion, pointers, structures, linked lists and files in C.
Data structures - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE 2014 till 2021• Number of teaching hours: 60.
• Number of students: 150.
• Content: Asymptotic analysis, ADTs, stacks, queues, trees, graphs, sorting, searching.
Advanced Algorithms and Competitive Programming - Lebanese University
BS 3RD YEAR COMPUTER SCIENCE - 2017 till 2021• Number of teaching hours: 40.
• Number of students: 20.
• Collaboration with Dr Siba Haidar
• Content: Complexity of recursive functions, complete search, divide and conquer, greedy algorithms, dynamic programming, advanced topics in graph, string processing, computational geometry
Programming with Python - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE 2020 till 2021• Number of teaching hours: 30.
• Number of students: 150.
Numerical in Python - ESIEA
4th YEAR ENGINEERING - 2022• Number of teaching hours: 18.
• Number of students: 30.
• Goal: Starting from a statement, the student can formulate the mathematical modeling of the problem, and implement the methods adapted to its optimization.
Image Processing in Python - ESIEA
4th YEAR ENGINEERING - 2022• Number of teaching hours: 12.
• Number of students: 45.
Machine Learning - ESIEA
4th YEAR ENGINEERING - 2022• Number of teaching hours: 12.
• Number of students: 24.
• Goal: Understand the principles and objectives of supervised learning from examples. Master the algorithms (nearest-neighbor, multilayer and convolution networks). Be able to size and optimize an algorithm on a training dataset and using a validation set. Measure the impact of parameters on performance and avoid overlearning.
Graph Theory and Applications - Lebanese University / ESIEA
BS 2ND YEAR COMPUTER SCIENCE - 2020 - 2021• Number of teaching hours: 30.
• Number of students: 120.
2ND YEAR ENGINEERING - 2021 - 2022
• Number of teaching hours: 48.
• Number of students: 28.
• Goal: The intended learning outcome of this course is to give good knowledge of graph theoretical concepts, and to practice how to use them in mathematics, natural science and computer science. After completing the course, students should be able to: know some important classes of graph theoretic problems; be able to formulate central theorems about trees, matching, connectivity, colouring and planar graphs; be able to describe and apply some basic algorithms for graphs; be able to use graph theory as a modelling tool.
Data Compression and Encoding - ESIEA
2ND YEAR ENGINEERING - 2021 - 2022• Number of teaching hours: 18.
• Number of students: 28.
• Goal: The prupose of this course is to: Acquire basic knowledge about the transport of any information; Know how the main data coding and compression techniques work; Design and implement compression algorithms adapted to a given problem
Introduction to logic and logic programming - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE - 2013 till 2015• Number of teaching hours: 30.
• Number of students: 120.
• Content: Propositional logic, Predicate logic, Prolog programming.
Advanced databases - Lebanese University
MASTER 1 COMPUTER SCIENCE - 2013 till 2018• Number of teaching hours: 48.
• Number of students: 30.
• Content: System crash, logs, transactions, concurrency, distributed databases, indexing, Btrees, B+trees, hashing, indexing, optimizing queries, deductive databases, object oriented databases. Oracle.
STRI PROFESSIONAL MASTER 1 - 2013 till 2015
• Number of teaching hours: 36.
• Number of students: 10.
• Content: PL/SQL principles, exceptions, cursors, triggers, collections, records, dynamic SQL, Bulk SQL, procedures, functions and packages
Knowledge Representation - Lebanese University
MASTER 1 COMPUTER SCIENCE - 2013 till 2017• Number of teaching hours: 36.
• Number of students: 20.
• Content: Predicates, conceptual graphs, ontology, knowledge Representation, web semantic.
Functional Programming - Lebanese University
BS 2ND YEAR COMPUTER SCIENCE - 2018 till 2021• Number of teaching hours: 30.
• Number of students: 50.
• Content: Python control statements, functions, strings, lists, dictionaries, tuples, sets, files, regular expressions, lambda expressions, iterators, generators and list comprehension.
Graphical interface and Applications - Lebanese University
BS 3RD YEAR COMPUTER SCIENCE - 2017-2018• Number of teaching hours: 36.
• Number of students: 50.
• Content: Design Patters: creational, structural and behavioral Patterns. JavaFX.
Server Side Web development - Lebanese University
BS 3RD YEAR COMPUTER SCIENCE - 2017-2018• Number of teaching hours: 50.
• Number of students: 50.
• Content: Web architecture. PHP language. PHP forms. Sessions. Connection with a DBMS (MySQL). Ajax and jQuery. Security(injections). Introduction to XML and XSLT. Regular expressions.
Distributed Applications programming - Lebanese University
MASTER 1 COMPUTER SCIENCE - 2012 till 2019• Number of teaching hours: 48.
• Number of students: 40.
• Content: Students learned to develop distributed applications. The course covered Java streams, TCP sockets, UDP datagram, Multithreading, RMI, serialization and agents.
Systems administration in networks - Lebanese University
STRI PROFESSIONAL MASTER 1 - 2013 till 2015• Number of teaching hours: 48.
• Number of students: 10.
• Content: Install, configure Windows servers. Configure server roles and features. Create and configure virtual machines. Deploy and configure DHCP and DNS services. Create and manage Active Directory users, groups and Organizational Units. Create Group Policy Objects. Configure Firewalls.
Pattern Recognition - Lebanese University
SIGNALS, TELECOMS, IMAGE AND SPEECH (STIP), FIELD ELECTROMECHANICAL ENGINEERING (FEME) AND CYBERSECURITY (CBRS) MASTER 2 - 2021• Number of teaching hours: 8.
• Number of students: 36.
• Collaboration with Dr Siba Haidar and Dr Yasser Mohanna.
s • Content: CNN Image Classification Lab. DNN and CNN Handwritten Digit Recognition Lab.
Scientific English - Lebanese University
GAME DESIGN AND DEVELOPMENT, CYBERSECURITY AND BIOSTATISTICS MASTER 2 - 2016-2017• Number of teaching hours: 36.
• Number of students: 30.
• Content: Planning, preparing and delivering a presentation. Presenting in conferences. Writing reports. Creating a CV and LinkedIn profile. Latex for papers, reports, CVs and presentations.
Initatiation to Scientific Research - ESIEA
To come.My Publications Publications
International Journal Articles
- Yaacoub, A.; Da Rugna, J. and Assaghir Z. (2024). Assessing AI-Generated Questions' Alignment with Cognitive Frameworks in Educational Assessment. In proceedings: 17th International Conference on Computer Science and Information Technology - ICCSIT 2024.
- Zainab Assaghir, Antoun Yaacoub, Sara Makki (2017). Network intrusion detection demystified using classification trees. In International Journal on Recent and Innovation Trends in Computing and Communication.
- Antoun Yaacoub, Zainab Assaghir (2017). Bottom-Up Bisimilar Datalog Goals. In American Journal of Engineering Research.
- Antoun Yaacoub, Hassan Safwan (2017). On the decidability of the information flow in Datalog using Magic Sets. In International Journal of Scientific Development and Research.
- Antoun Yaacoub (2016). Bisimilarity, Datalog And Negation. In International Journal of Computer Science: Theory and Application.
- Antoun Yaacoub, Ali Awada (2016). Information Flow In Datalog Using Very Naive And Semi Naive Bottom-Up Evaluation Techniques. In International Journal of Computer Science: Theory and Application.
- Antoun Yaacoub, Ali Awada (2015). Inference Control On Information Flow In Logic Programming. In International Journal of Computer Science: Theory and Application.
- Antoun Yaacoub, Ali Awada, Habib Kobeissi (2015). Information Flow in Concurrent Logic Programming. In British Journal of Mathematics and Computer Science.
- Antoun Yaacoub (2012). Towards an information flow in logic programming. In International Journal of Computer Science Issues.
- Ali Awada, May Dehayni, Antoun Yaacoub (2011). An ATMS-Based Tool for Locating Honor Cards in Rubber Bridge. In Journal of Emerging Trends in Computing and Information Sciences.
- Ali Awada, May Dehayni, Antoun Yaacoub (2011). Visualizing relationships in Hierarchical Small World Networks. In International Journal of Computer Science and Network Security.
International Conference Articles
- Laborde, S., Cousseau, M., Yaacoub, A., and Prevost, L. (2025). Semantic Retention and Extreme Compression in LLMs. In Proc. International Joint Conference on Neural Networks...
- Yaacoub, A., Tarnpradab, S., Khumprom, P., Assaghir, Z., Prevost, L., and Da-Rugna, J. (2025). Enhancing AI-Driven Education: Integrating Cognitive Frameworks, Linguistic Feedback Analysis, and Ethical Considerations for Improved Content Generation. In Proc. International Joint Conference on Neural Networks...
- Yaacoub, A., Assaghir, Z., Prevost, L., and Da-Rugna, J. (2025). Analyzing Feedback Mechanisms in AI-Generated MCQs: Insights into Readability, Lexical Properties, and Levels of Challenge. In Proc. 2025 9th Int. Conf. on Computer, Software and Modeling...
- Yaacoub, A., Assaghir, Z., and Da-Rugna, J. (2025). Cognitive Depth Enhancement in AI-Driven Educational Tools via SOLO Taxonomy. In Proc. Third International Conference on Advances in Computing Research...
- Al Zein, H.; Yaacoub, A. and Haidar, S. (2024). Beehive State and Events Recognition Through Sound Analysis Using TinyML. In proceedings: 17th International Conference on Machine Vision (ICMV 2024).
- Desplanche, M.; Haidar, S.; Yaacoub, A. and Barrois, B. (2024). Protecting Beehive Health: Varroa Mite Detection Using Gen-AI. In proceedings: 17th International Conference on Machine Vision (ICMV 2024).
- Yaacoub, A.; Haidar, S. and Da Rugna, J. (2024). OneClickQuiz: Instant GEN AI-Driven Quiz Generation in Moodle. In proceedings: Conference on Sustainable Energy Education – SEED 2024.
- Quentin Pajon, Swan Serre, Hugo Wissocq, Léo Rabaud, Siba Haidar, Antoun Yaacoub (2024). Balancing Accuracy and Training Time in Federated Learning for Violence Detection in Surveillance Videos: A Study of Neural Network Architectures. Journal of Computer Science and Technology.
- Quentin Pajon, Swan Serre, Hugo Wissocq, Léo Rabaud, Siba Haidar, Antoun Yaacoub (2023). Balancing Accuracy and Training Time in Federated Learning for Violence Detection in Surveillance Videos: A Study of Neural Network Architectures. In International Workshop on Trustworthy Federated Learning in Conjunction with IJCAI 2023 (FL-IJCAI'23), Macau.
- Siba Haidar, Antoun Yaacoub, Felicia Ionascu (2023). Evaluating the Effectiveness of a New Programming Teaching Methodology Using CodeRunner. In The Learning Ideas Conference 2023, 16th annual conference, New York.
- Antoun Yaacoub, Zainab Assaghir, Sara Makki, Radwan Almokdad (2019). Diagnosing Clinical Manifestation of Apathy Using Machine Learning and Micro-facial Expressions Detection. In International Symposium on Computer Science and Intelligent Control (ISCSIC), Netherlands.
- Philippe Balbiani, Antoun Yaacoub (2012). Deciding the bisimilarity relation between Datalog goals (regular paper). In European Conference on Logics in Artificial Intelligence (JELIA 2012), Toulouse, France.
Conference Articles (No Published Proceedings)
Posters
- Edouard Nadaud, Thibault Geoffroy, Tesnim Khelifi, Antoun Yaacoub, Siba Haidar, Nourhene Ban Rabah, Jean-Pierre Aubin (2023). From Learning Management System to Affective Tutoring system: a preliminary study. In International Workshop on Multimodal, Interactive Interfaces for Education, part of the 25th ACM International Conference on Multimodal Interaction (ICMI 2023), Paris.
- Kassir, M., Haidar, S., & Yaacoub, A. (2025). Exploring Personalized Federated Learning Architectures for Violence Detection in Surveillance Videos. arXiv preprint arXiv:2504.00857.
Thesis and HDR
Semantic Retention and Extreme Compression in LLMs
This paper presents cutting-edge insights into semantic retention and extreme compression techniques for Large Language Models (LLMs), crucial for building efficient and scalable next-generation AI models. This work explores advanced methods applicable to optimizing LLM performance in real-world scenarios, particularly relevant for large-scale AI deployment and optimization.
Enhancing AI-Driven Education: Integrating Cognitive Frameworks, Linguistic Feedback Analysis, and Ethical Considerations for Improved Content Generation
This research demonstrates how to leverage Generative AI for advanced content creation in educational contexts, integrating cognitive frameworks and linguistic feedback. A core focus is on the ethical implications of AI development, making it highly relevant for responsible AI system design and deployment in real-world applications.
Analyzing Feedback Mechanisms in AI-Generated MCQs: Insights into Readability, Lexical Properties, and Levels of Challenge
This paper explores the analytical capabilities of LLMs to understand and extract key features from AI-generated content, focusing on readability and lexical properties of Multiple Choice Questions (MCQs). This work is crucial for developing sophisticated AI-driven assessment tools and for understanding model behavior and output in practical learning contexts. It highlights aspects of NLP and model interpretability.
Cognitive Depth Enhancement in AI-Driven Educational Tools via SOLO Taxonomy
This research investigates methods for enhancing the cognitive depth of AI-driven educational tools using the SOLO Taxonomy. It explores advanced AI techniques and frameworks to generate content that promotes deeper learning and critical thinking, with direct applications in adaptive educational systems and intelligent tutoring.
Assessing AI-Generated Questions' Alignment with Cognitive Frameworks in Educational Assessment
This paper focuses on the evaluation and alignment of AI-generated questions, particularly those from LLMs, with established cognitive frameworks in educational assessment. It highlights the importance of robust evaluation methods for AI-driven content generation, ensuring quality and pedagogical effectiveness in intelligent learning systems.
Beehive State and Events Recognition Through Sound Analysis Using TinyML
This research explores the application of TinyML for real-time detection and recognition of specific states and events within beehives through sound analysis. It demonstrates practical skills in deploying efficient machine learning models on resource-constrained devices for predictive analytics and anomaly detection.
Protecting Beehive Health: Varroa Mite Detection Using Gen-AI
This paper demonstrates the practical application of Generative AI for real-world detection challenges, specifically in identifying threats to beehive health. It showcases interdisciplinary research and the use of cutting-edge AI technologies for practical solutions in complex environmental monitoring.
Balancing Accuracy and Training Time in Federated Learning for Violence Detection in Surveillance Videos: A Study of Neural Network Architectures
This research provides experience with large-scale model optimization and architecture design within a federated learning context. It's crucial for developing robust AI solutions for production environments, addressing challenges of privacy and distributed training for complex analytical tasks like anomaly detection in video surveillance.
Diagnosing Clinical Manifestation of Apathy Using Machine Learning and Micro-facial Expressions Detection
This study explores the application of Machine Learning and micro-facial expression detection systems for diagnosing clinical manifestations of apathy. It involves applying Histogram of Oriented Gradients (HOG) as a feature descriptor on video datasets and utilizing statistical models (Decision Trees, Logistic Regression) to identify influencing factors. This work is relevant for applied ML research in predictive analytics and understanding complex patterns in behavioral data.
Network Intrusion Detection Demystified Using Classification Trees
This research focuses on applying Classification And Regression Tree (CART) algorithms to detect network intrusions in enterprise environments. It demonstrates expertise in supervised learning techniques for security applications and classification problems, achieving high accuracy with low false-negative and false-positive rates for critical systems.
Bottom-Up Bisimilar Datalog Goals
This paper extends concepts of bisimilarity relation in Datalog programs using a bottom-up evaluation approach via Magic Sets. It proves the decidability of information flow and analyzes its computational complexity, relevant for logical AI, knowledge graphs, and formal verification in AI systems.
On the Decidability of the Information Flow in Datalog Using Magic Sets
This research investigates the decidability of information flow in Datalog programs using the Magic Sets bottom-up evaluation approach. It proves the existence of information flow is decidable with EXPTIME-complete complexity, contributing to formal methods in AI and secure system design.
Bisimilarity, Datalog And Negation
This paper extends the concept of bisimilarity relation between Datalog goals from positive Datalog programs to stratified and restricted programs with negation. It addresses the decidability of bisimilarity, enhancing formal methods for reasoning in complex AI systems and knowledge representation.
Information Flow In Datalog Using Very Naive And Semi Naive Bottom-Up Evaluation Techniques
This work formalizes information flow detection in Datalog using Very Naive and Semi Naive bottom-up evaluation algorithms. It proves decidability and analyzes computational complexities, contributing to security and traceability in knowledge-based systems and AI applications.
Inference Control On Information Flow In Logic Programming
This paper proposes a formal representation of inference control on information flow theory in logic programming. It introduces notions of indistinguishability, protection mechanisms, and confidentiality policies for controlling sensitive information in deductive databases and AI systems, relevant for ethical AI and data privacy.
Information Flow in Concurrent Logic Programming
This research presents a formalization of information flow detection in concurrent logic programming and applies it to deadlock detection. It defines information flow based on observation and transition systems, contributing to the development of robust and secure concurrent AI systems.
Towards an Information Flow in Logic Programming
This paper provides a theoretical foundation for information flow in logic programming, comparing various definitions based on success/failure, substitution answers, and bisimulation. It addresses decision procedures and complexity for specific classes of logic programs, relevant for foundational AI and secure data handling.
An ATMS-Based Tool for Locating Honor Cards in Rubber Bridge
This research proposes an AI-based approach for assisting Bridge players in locating honor cards, utilizing abduction reasoning and an Assumption-based Truth Maintenance System (ATMS). It demonstrates expertise in AI reasoning, knowledge representation, and constraint satisfaction for complex problem-solving.
Visualizing Relationships in Hierarchical Small World Networks
This research presents a method for studying and visualizing relationships in hierarchical small world networks (HSWN) applied to various domains like dictionaries and web pages. It develops an approach to quantify relationships using Markovian matrices and analyze large graph structures, relevant for data visualization, network analysis, and knowledge graph construction.
OneClickQuiz: Instant GEN AI-Driven Quiz Generation in Moodle
This paper presents an innovative strategy for generating quizzes and multiple-choice questions (MCQs) within Moodle, leveraging advanced capabilities of Generative AI (Google's PALM2 models). It focuses on streamlining educational workflows, enhancing academic engagement, and demonstrating the transformative potential of integrating cutting-edge AI technologies into educational platforms, especially for personalized learning.
Evaluating the Effectiveness of a New Programming Teaching Methodology Using CodeRunner
This research evaluates an innovative teaching approach for programming courses combining online coding platforms, automated grading, interactive visualizations, and offline materials. It demonstrates enhanced student engagement and learning outcomes through modern techno-pedagogical methods, highlighting the impact of technology on informatics education.
Deciding the Bisimilarity Relation Between Datalog Goals (Regular Paper)
This paper introduces the concept of bisimulation between Datalog goals, where two goals are bisimilar if their SLD-trees are bisimilar. It addresses the decidability problem for given stratified or restricted Datalog programs, proving decidability in 2EXPTIME. This contributes to formal methods in AI, logic programming, and knowledge representation systems.
Balancing Accuracy and Training Time in Federated Learning for Violence Detection in Surveillance Videos: A Study of Neural Network Architectures
This study investigates machine learning techniques for violence detection in videos within a federated learning context. It includes experiments with spatio-temporal features, architectural comparisons, and adaptation of centralized datasets. It demonstrates expertise in large-scale model optimization, distributed training, and robust AI solutions for complex real-world applications.
Information Flow in Logic Programming (Journées d’Intelligence Artificielle Fondamentale)
This paper proposes a theoretical foundation for information flow in logic programming, comparing various definitions and analyzing their decision procedures and complexity for specific classes of logic programs. It contributes to fundamental AI research, logic-based reasoning, and secure data handling.
From Learning Management System to Affective Tutoring System: A Preliminary Study
This study investigates the combination of indicators (performance, behavioral/emotional engagement) to identify students experiencing difficulties, using LMS digital traces and webcam images for emotional analysis. It explores the role of emotions in academic outcomes and contributes to the development of intelligent tutoring systems and AI-driven personalized learning.
Information Flow in Logic Programming (PhD Thesis)
This doctoral thesis addresses information flow in logic programming, proposing theoretical foundations, various flow definitions (success/failure, substitution answers, bisimulation), and decision procedures/complexity for specific program classes. It also introduces the concept of bisimulation between Datalog goals and proposes a secure, precise security mechanism for deductive databases, contributing to foundational AI, logic-based reasoning, and secure data management.
Get in Touch Contact
Message Me
Contact Info
If you have any questions, comments, or if you simply want to talk please contact me! For instant replies, follow me on Instagram, Facebook, Twitter, ...