Future Generation – Towards a Personal Expert System

While experimental and commercial specialized expert systems have been created for several decades there has been no examination of the requirements associated with an expert system that is tailored to an individual user.  A more comprehensive assemblage entails an adaptation of a rule-based expert system architecture in which the facts, rules, and inference capabilities reside on the client device.  A personal expert system (PES) of this type would be dedicated to:

- Learning the user's preferences.

- Increasing in knowledge through interactions with the operator.

- Acquiring facts from the external environment.

- Applying knowledge libraries made available by businesses and organizations.

The PES would enhance the user’s decision-making ability by learning the user’s inclinations and providing focused recommendations in specialized knowledge areas, further enhanced by feedback from information retrieval.  With the user’s consent commercial vendors and other organizations would provide properly formatted content through a controlled gateway to expand the system’s knowledge base.  The seminal knowledge core would enlarge according to the dynamics of the user, ameliorating the user’s communication with the world in ways distinctively beneficial to the user.  The PES would dynamically evolve within the confines managed by the user.  This approach is contingent upon common data format and ontology standards among participants.

Such a personal expert system would interface both autonomously within parameters and under user direction with the external environment, over time becoming increasingly familiar with the user’s preferences.  It would provide the user’s preferences to external processes as permitted and appropriate.  It would also capture the user’s human behavior and recognize its evolution over time.  The PES could essentially represent the user as a virtual presence based upon identity, status, location, and other predefined preference parameters.  This intelligent learning system would in effect become the user’s enduring companion, retrieving and providing information from external sources, interacting with ambient sensors for identification and adjustments, and offering recommendations based upon past experiences and current data as shown in the illustration below.  A PES, transferrable among devices, could become a pivot point around which future interaction between the owner and society would be individualized.