Pervasive-based Modeling And Application Of Smart Space

With the computing technology developing, especially after the proposal of pervasive computing, intelligent human-machine interaction gradually begins to prevail. Pervasive computing in smart space mainly focuses on running tasks other than actual computing devices. In some way, users wouldn't be aware of their existence. This kind of quality requires applications not to take the input or obvious commands as their triggering points, but to make the decisions depending on the already known history and information intelligently. Based on the above, in order to better implement the idea of"transparency"in pervasive computing, integrating interactive context-aware module within smart space becomes one key area worth researching on.In this paper, we first introduce three new rising concepts, pervasive computing, smart space and context-aware. Then based on the above three technologies, we bring on the framework named Generic Pervasive-Based Intelligent Space. This framework combines physical space with human attention tightly. Its significances include offering a unique and adoptable framework, highlighting out one abstract conceptual model, and laying the ground for realistic pervasive system's construction.Next, based on the framework of GPBIS, we construct a real system named Shadow-Like Intelligent Tracking Pervasive Campus. The main innovations of SLITPS are: (1) We propose a lightweight user-shadow model to cater for the context awareness and track the user anytime, anywhere. Moreover, the user-shadow can be tailored to the context and user preferences. 2) We provide scalable, distributed resource discovery mechanism, which divides resources into two layers. The outer layer spies on the joining and leaving of resource nodes, and notifies the inner layer timely. Therefore, the nodes basically have no need to update status lists by themselves. 3) We offer a potent context inference mechanism. SLITPC employs the ontology to represent context and Bayesian networks for causal reasoning, which offers great help to certain and uncertain reasoning.In order to implement the transparency of pervasive computing, we add collaborative filtering system into SLITPC, which serves as a music recommendation module for users. The innovations includes: (1) We adopt"song and artist"double-criteria recommendation strategy. (2) We develop an implicit rating extraction mechanism. (3) We develop a Centralized recommendation and a Gossip-based P2P recommendation, and integrate these two algorithms together to offer better services to users. (4) We have shed lights on the serendipity problem that is common in most music recommendation systems.