I'm doing research and the incorporation of a knowledge base into inductive machine learning, and proposes combining prior domain knowledge in three key issues of inductive machine learning, namely consistency, generalization and convergence. The variety and diversity of the domain knowledge led to difficulties in measuring the relationship between domain knowledge and tasks and represent the learning system. with no universal solution to this problem. I want you to help me in making the theory and applications to solve this problem, the application must be object-oriented, so that one day easily developed.
Please do not disturb if you do not have any knowledge about this!