Existing large and shareable repositories of knowledge, e.g., libraries, are dependent on the notion of context for their structure and efficient use. Books are classified and catalogued into subject areas, authors, keyword areas, etc. Each of these arrangements facilitate the appropriate access to, and hence use of, relevant parts of the repository. This paper argues that the notion of context,
... [Show full abstract] although complex and little understood in the AI community, must be one of the intrinsic parameters of any large and shareable knowledge-base system. It introduces some details of context as found in the literature to highlight its complexity and how little understood the notion of context is. Electronic document systems is suggested as a most promising area for researching into context applicable within large and shareable knowledge-base systems.