Scenario analysis using Bayesian networks: A case study in energy sector
Knowledge-Based Systems
| Hi-index | 0.00 |
Scenarios are narratives that illustrate future possibilities or existing systems, and help policy makers and system designers choose among alternative courses of action. Scenario-based decision-making crosses many domains and multiple perspectives. Domain-specific techniques for encoding, simulating, and manipulating scenarios exist, however there is no scenario representation capable of supporting the wide spectrum of formality from executable simulation to free-form text to streaming media descriptions. The claim of this research is that there is a computer readable scenario framework that can capture the semantics of a problem domain and make scenarios an active part of decision making. This dissertation describes a scenario ontology derived by examining alternate forms of narrative: thought experiments, mental models, case-based reasoning, use cases, design patterns, screenwriting, film editing, intelligent agents, and other narrative domains. The scenario conceptual model was based on an analysis of forms of narrative and the activities of storytelling. This method separates what a narrative is from how it is used. The research contribution is the development of the Hyperscenario Framework. A hyperscenario is a scenario representation containing link structures for navigation between scenario elements. The Hyperscenario Framework consists of the scenario ontology, scenario grammar, and a scenario specification called Scenario Markup Language (SCML). The results of the web-enabled simulation experiment validate the improvement on decision-making due to the Hyperscenario Framework.