The learning process for a user becomes seriously restrictive in trying to discover the relationships between concepts and in searching for a part of concept (called an object) such as a solved example illustrating the concept or the application of the concept. This article presents the theory for building a relational graph that depicts how concepts are linked to each other. By selecting to zoom in on a particular concept, the user's view changes to conceptual graph where s/he can view and access all the objects related to a particular concept. Rule-based algorithms are presented to identify objects of a concept, to determine concept boundary, and to build the trees. The lecture material on Algorithms course from MIT is used for experimentation of the ideas. In addition to efficient searching for a desired topic, the system also enhances the understanding and the learning of the user.