Rule Selection for Collaborative Ubiquitous Smart Device Development: Rough Set Based Approach

摘要

Comparing with general mobile devices, Ubiquitous Smart Device (USD) is characterized by its capability to generate or use context data for autonomous services, and it provides users with personalized and situation-aware interfaces. While the USD development requires more knowledge-intensive and collaborative environment, the capture, retrieval, accessibility, and reusability of that design knowledge are increasingly critical. In the design collaboration, the cumulative, evolutionary design information and design rules behind the USD design are infrequently captured and often difficult to hurdle due to its complexity. Rough set theory synthesizes approximation of concepts, analyzes data by discovering patterns, and classifies into certain decision classes. Such patterns can be extracted from data by means of methods based on Boolean reasoning and discernibility. In this paper, a rough set theory generates demanded rules and selects the appropriate minimal rules among the demanded rules associated to USD physical component design. The presented method shows the feasibility of rough-set based rule selection considering complex design data objects of USD physical components.