Abstract
Although the introduction of powerful hypertext authoring tools simplifies the task of creating medical “hyperbooks”, these tools do not greatly simplify the difficult task of creating and using “hyperlibraries” composed of collections of “hyperbooks”. We argue that in addition to reintroducing many of the classical problems of full-text document retrieval, hypertext's critical emphasis on document structure and link semantics appears to introduce a newer class of information retrieval problems. We suggest that probabilistic and connectionist approaches to information retrieval might be applicable to medical hypertext.
Full text
PDF
