Knowledge representation contributions

  Although not the focus of my work, there are several contributions found in this work in the area of knowledge representation. First,   A FUNCTIONAL REPRESENTATION ALLOWS A CONTINUUM OF NOVEL CONCEPTS TO BE REPRESENTED IN A FLEXIBLE FASHION. By representing the function of a concept along with the primary and secondary attributes in relation to that function, a range of concepts are possible. I supported this claim in Chapter 4 by presenting the details of the representation and by arguing the validity of it, by appealing to both prior work in knowledge representation and by evaluating the model which implements that theory of knowledge (as shown in Chapter 8).

In addition to the basic knowledge representation issues, my theory makes use of a specific   ontology consisting of twenty top-level categories. I claimed that   FIVE ONTOLOGICAL DOMAINS AND FOUR ONTOLOGICAL TYPES ARE SUFFICIENT TO EXPLAIN THE WORLD AT A DEPTH WHICH ALLOWS A WIDE RANGE OF CONCEPTS TO BE REPRESENTED AND MANIPULATED. These ontological categories were shown in Chapter 4 in Figure 5. The claim was supported (also in Chapter 4) by appealing to a body of developmental psychology literature which discusses the various ontological divisions which arise as childhood development takes place. Added to this was a discussion demonstrating how various world entities would be represented in the scheme, as well as demonstrating the   ISAAC system making use of the   ontology during the course of reading (as seen in Chapter 7).

These two lowest claims together provide support for the claim that   CONCEPTS NEED TO BE REPRESENTED AND ORGANIZED FOR EFFICIENT UNDERSTANDING. This may seem like an obvious conclusion; however, I have presented the specific details of one method which shows itself to be sufficient for accomplishing this. If I discard the   ontology, some level of understanding is still possible; however, the effort required to arrive at this understanding is greatly increased. If no internal representation was allowed at all, then no internal manipulation of the concepts could occur, at least not in the symbolic paradigm in which my work exists. Thus, I can conclude that my functional representation and the     ontology I have developed work together in the reasoner to aid the understanding process.