Recent researches suggest the new model of adaptive learning that considers the interpersonal differences between learners when presenting a certain lesson for each individual student.  However, teachers are resisting the new model under the pretext of not wanting to rebuild their lessons again to follow the recommendations of this new model.  Therefore, the objective of this research is to overcome this resistance by automating the process of constructing adaptive lessons given a certain textbook. Therefore, this research follows a 2-step process: 1) identification and extraction of learning objects from the textbook, 2) composing the lessons using those extracted objects. Therefore, this research developed a YOLOv5 deep learning model, for book's elements identification, using a dataset of about 70 books, references and school textbooks, containing over 6000 pages in both English and Arabic languages of various specializations and knowledge domains.  The developed model has an accuracy of about 92%.  Those identified learning objects are then compiled into lessons with the aid of a concept ontology that is automatically constructed from the book's table of contents.
 
Special Issue of AEIC 2024 (Electrical and System & Computer Engineering  Session)