Abstract
The creation of homogeneous student groups is a vital element to the learning process in Universities. Students usually collaborate and work in teams to gain insights, learn from their peers and improve their intellectual abilities. However, forming effective learning groups is a not a trivial activity especially that students vary according to multifold aspects like learning styles, performance, personality, commandment of language. As such differences increase the difficulty of grouping students effectively. This paper suggests an agent-based computational algorithm that aims to achieve the automatic formation of effective groups in a learning environment considering the unique characteristics and differences of university students. To achieve this goal, we propose a two phase research process, wherein the first phase a survey is administered and distributed to university students and instructors to collect information about student learning characteristics and any potential constraints, and in the second phase, these requirements are combined and translated into a smart agent-based architecture by leveraging the advantages of the multi-agent paradigm. The purported architecture will have the capacity to be customized to accommodate the specific needs of course instructors within various educational institutions and contexts.