TY - JOUR
T1 - Formal real-time imagination
AU - Nirkhe, Madhura
PY - 1995
Y1 - 1995
N2 - Formal real-time imagination is a term that may curiously describe the activities of a commonsense agent in a real-time setting in general, and in a tight deadline situation in particular. We briefly describe an active-logic mechanism that fits this description. Temporal projection is an essential component of realtime planning. We draw a parallel between imagination as we understand it in human context and the capacity of the automated agent toformulate mental images of possible scenarios and plans of action in the course of its reasoning. We outline a treatment of temporal issues of significance to a time-situated reasoning mechanism in a dynamic setting with deadlines. The Yale shooting problem is a benchmark problem in temporal reasoning. We demonstrate how the active-logic planning mechanism successfully handles some interesting real-time variants of the Yale shooting problem. The solutions to each of these illustrate the agents ability to form contexts within which to reason, to project in each context thus formed by applying default inferences, and to revise and extend its conclusions within each context by applying time-sensitive inference rules, and most importantly, to account forall the time spent in the process.
AB - Formal real-time imagination is a term that may curiously describe the activities of a commonsense agent in a real-time setting in general, and in a tight deadline situation in particular. We briefly describe an active-logic mechanism that fits this description. Temporal projection is an essential component of realtime planning. We draw a parallel between imagination as we understand it in human context and the capacity of the automated agent toformulate mental images of possible scenarios and plans of action in the course of its reasoning. We outline a treatment of temporal issues of significance to a time-situated reasoning mechanism in a dynamic setting with deadlines. The Yale shooting problem is a benchmark problem in temporal reasoning. We demonstrate how the active-logic planning mechanism successfully handles some interesting real-time variants of the Yale shooting problem. The solutions to each of these illustrate the agents ability to form contexts within which to reason, to project in each context thus formed by applying default inferences, and to revise and extend its conclusions within each context by applying time-sensitive inference rules, and most importantly, to account forall the time spent in the process.
UR - http://www.scopus.com/inward/record.url?scp=0029325312&partnerID=8YFLogxK
U2 - 10.3233/FI-1995-232348
DO - 10.3233/FI-1995-232348
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
SN - 0169-2968
VL - 23
SP - 371
EP - 394
JO - Fundamenta Informaticae
JF - Fundamenta Informaticae
IS - 2-4
ER -