主观Bayes网络在普适计算中的应用研究

Application Research and on Subjective Bayesian Network in Pervasive Computing

由于硬件系统的异构性和无线网络的相对不稳定,造成直接由传感器获得的上下文的不确定性和不一致性。笔者针对此问题把主观Bayes网络模型应用到普适环境中对上下文进行推理,并且对OWL进行了扩展——SBOWL语言,以适应主观Bayes网络模型在上下文推理中的应用。通过在原型系统CIS的上下文推理层应用主观Bayes网络,验证了该方法的可行性。

Context-aware is one of the most important technologies in pervasive computing. Due to heterogeneity of hardware structure and instability of wireless network, the context directly obtained from sensors is often uncertain and inconsistent. This paper applies Subjective Bayesian Network to infer about context in pervasive environment and proposes SBOWL language, which extended OWL, to make Subjective Bayesian Network fit for infering about context. The feasibility of our method is proved by applying it to the context reasoning layer of CIS prototype system.