A Simple-Minded Network Model with Context-like Objects - Bruce Edmonds
Perhaps a more natural way of formalising context in order to compare different conceptions, is to not distinguish contexts from other object of reasoning, learning etc. but to be chosen such that certain object act like contexts in certain circumstances. Thus one would like a reasonably simple formalism such that particular kinds of model of context and context-related structures can be seen as special cases of the more general structure. I put forward one candidate below - I have suspicions that it is too simple for the job, but it is better to start simply and elaborate rather than start with an over-complex formalism that may obscure the essential details and hence the debate.
Generated with CERN WebMaker