A new domain-independent knowledge-based inference structure is presen
ted, specific to the task of abstracting higher-level concepts from ti
me-stamped data. The framework includes a model of time, parameters, e
vents, and contexts. A formal specification of a domain's temporal abs
traction knowledge supports acquisition, maintenance, reuse, and shari
ng of that knowledge. The knowledge-based temporal abstraction method
decomposes the temporal abstraction task into five subtasks. These sub
tasks are solved by five domain-independent temporal abstraction mecha
nisms. The temporal abstraction mechanisms depend on four domain-speci
fic knowledge types: structural, classification (functional), temporal
semantic (logical), and temporal dynamic (probabilistic) knowledge. D
omain values for all knowledge types are specified when a temporal abs
traction system is developed, The knowledge-based temporal abstraction
method has been implemented in the RESUME system, and has been evalua
ted in several clinical domains (protocol-based care, monitoring of ch
ildren's growth, and therapy of diabetes) and in an engineering domain
(monitoring of traffic control), with encouraging results. (C) 1997 E
lsevier Science B.V.