Empowering Education: AMCQA Generation with T5 Transformers
Radhwan Hussein Abdulzhraa Al Sagheer · Zenodo (CERN European Organization for Nuclear Research) · 2023
The growing popularity of e-learning kits and the increasing number of online knowledge seekers in educational institutions such as universities, schools, childcare institutions and other scientific organizations have presented many challenges and difficulties for educators who are interested in creating modern tests. This makes fair and effective assessment of students difficult, and a major obstacle is the manual formulation of test-style questions, which disrupts the smooth integration of educational activities and hinders the development of innovative teaching methods based on diverse questions. Therefore, there is an urgent need to explore modern solutions to help teachers overcome these difficulties and ensure a smooth testing process, accurate answers and fair assessment of students. This paper introduces Empowerment Education (EE), a model specifically designed for automatic answering multiple choice questions (AMCQAs) from real-life texts. To overcome the challenges and obstacles that hinder the effective management of sequential tests, the model uses advanced natural language processing and machine learning tools based on the T5 architecture. This system aims to simplify the generation of various contextually relevant questions to enhance students' knowledge and critical thinking abilities. In addition to being highly suitable for educational settings, it can have practical applications in an industrial context, facilitating knowledge-sharing and dissemination.