Research And Implementation Of Context-Aware Technology In Wot Environments

Context-aware technology is an important part of intelligent interaction field in computer science. With the rapid development of the Internet of things, human, physical space and information technology are integrated greatly in real life. As a result, context-aware technology have been widely used and popularized in many fields such as smart hospital, smart home and smart grid, etc. In recent years, the concept of WoT (Web of Things), which applied web technology to Internet of Things have put forward new requirements and challenges for context-aware technology. On one hand, various sensors and devices have been accessed from the underlying layer, which have different resources’models and characteristics. The inconsistency of data structure has led to the difficulties of information sharing in different domains of WoT. On the other hand, the real-time and massive WoT data also has requested more on context-aware reasoning technology. How to realize the fast inference of massive multi-source and heterogeneous data and improve the inference time efficiency is a key issue for context-aware technology. Finally, based on the openness of WoT environments, we have to find out an effective set of architecture and mechanism to realize the effective management and rapid reactions to environment changes.This paper focuses on the research of context-aware key technology in WoT environments. Firstly, we give a detailed analysis of the key context-aware technology in WoT environments such as context-aware system, context modeling and context reasoning, put forward the shortcomings and defects of existing technology and summarize the new challenges and requirements for WoT environments. On the basis of this analysis we design the Uniform Context-aware Middleware Architecture (UCAM) for WoT environments, which realize the context acquisition, context modeling, context storage, context reasoning and context service opening. UCAM supports two context processing mechanism:passive request/response from top to bottom and initiative subscription/notification from bottom to top. Then we study on two key technologies in context-aware technology:context modeling and context reasoning. Context modeling is the base of the context-aware technology, which solves the problem of context description and representation. In this paper, we propose the Resource-Oriented Context Model (ROCM) on the basis of OWL-S ontology model, which aims at the web service description. ROCM provides the description and representation of WoT elements such as user, resource, and service and supports the context reasoning in certain and uncertain ways. Context reasoning includes Semantic Web Language Reasoning (SWRL) and Bayesian network reasoning. Considering the requirements of WoT environments this paper improves the core Rete algorithm of context reasoning which makes context reasoning engine can better adapt to the environment changes of physical spaces and improve the time efficiency of context reasoning.