Abstract—Comprehending
Object-Oriented
Programming
(OOP) concepts is a difficult task especially for novice students.
This usually happens during the transition form learning
fundamental concepts to object-oriented (OO) concepts. When
given an OO problem to solve, novices find it hard to relate with
objects. If novices can view the world based on real objects, this
can help them solve their problem of comprehending the OO
concepts. In this paper, we propose to design an agent model to
understand the semantics of OO Java source codes. The agents
are designed based on the Belief-Desire-Intention (BDI)
architecture. Three agents namely GUI agent, semantic agent
and novice agents are constructed. The GUI agent is controlled
by the user to provide source codes examples. The semantic agent
submits the source codes to the novice agent and explains the
semantic or meaning of the source codes to the novice agent. The
process involves source codes comparison technique. This model
provides an OOP semantics knowledge representation based on
intelligent agents.