Throughout the industrial era societal advancement could be attributed in large part to introduction a plethora of electromechanical machines all of which exploited a key concept known as Mechanical Advantage. In the post-industrial era exploitation of knowledge is emerging as the key enabler for societal advancement. With the advent of the Internet and the Web, while there is no dearth of knowledge, what is lacking is an efficient and practical mechanism for organizing knowledge and presenting it in a comprehensible form appropriate for every context. This is the fundamental problem addressed by my dissertation. We begin by proposing a novel architecture for creating a Knowledge Advantage Machine (KaM), one which enables a knowledge worker to bring to bear a larger amount of knowledge to solve a problem in a shorter time. This is analogous to an electromechanical machine that enables an industrial worker to bring to bear a large amount of power to perform a task thus improving worker productivity. This work is based on the premise that while a universal KaM is beyond the realm of possibility, a KaM specific to a particular type of knowledge worker is realizable because of the limited scope of his/her personal ontology used to organize all relevant knowledge objects. The proposed architecture is based on a “society of intelligent agents" which collaboratively discover, markup, and organize relevant knowledge objects into a semantic knowledge network on a continuing basis. This in-turn is exploited by another agent known as the Context Agent which determines the current context of the knowledge worker and makes available in a suitable form the relevant portion of the semantic network. In this dissertation we demonstrate the viability and extensibility of this architecture by building a prototype KaM for one type of knowledge worker such as a professor.
