The Thick ICE project requires an expert system to help ‘diagnose’ a patient when the symptoms do not stand out obviously to the user. The expert system takes information about a particular case within the domain of the system, and produces a qualitative result. The expert system would require a knowledge base, which is essentially a set of rules describing the relations between elements in the domain of knowledge. (i.e. condition(s) –> outcome(s)).
Due to the Android SDK being used for development, Java is widely supported and will be integrated into the app development. There are some open source java expert systems available such as JBoss and Jess; however the current open source expert systems are not directly compatible with integration with an android app. Jess is not supported at all, and also requires approval and payment of the source to use in the app’s development. JBoss is free to use and suitable, however it is a web-based application which would require RESTFUL services to access it through the android app. Although JBoss is available and possible to implement, it is not our first preference due to it being rather difficult to implement.
Another approach to the expert system is to manually code it ourselves through switch cases and a knowledge base. This provides the system with a IF (condition(s)) THEN (outcome(s)) structure which is used in the open source systems. This approach will be researched in greater detail.