泛在计算环境中社会化驱动的情境感知个性化信息服务研究

It is required providing personalized information service (PIS) at anytime and anywhere in ubiquitous computing environment (UCE). So massive, diverse and dynamic changing context information should be defined, provided, managed and utilized as context-aware is adopted to achieve PIS in that context-aware has been recognized an effective mean to achieve PIS. But, current most approaches for context-aware personalized information service (CAPIS) are depended on small experts and predefined. So, they cannot deal with the defining, providing, managing and utilizing of massive, diverse and dynamic changing context information in UCE. Therefore, a new idea is proposed which achieving CAPIS in UCE by social driven mode, and an approach for CAPIS is present which is depended on large users and dynamic and open. After theories about principles, collaborative mechanisms and processes for social driven CAPIS are studied, the affair-matter element model, the incidence space model and the complex supernetwork model for context and information resources are built, and the dynamic optimization method based on self-classifying, self-identifying and self-clustering for context and information resources is put forward. Based on these, the context discovery and retrieval method based on combination of incidence expanding, network discovering and optimal evaluating and the information optimizing-integrated discovery and retrieval method based combination of match discovering and optimal evaluating are proposed. Finally, the prototype system is built and the evaluation and application are studied. This study is driven by the requirement from application and gives a new idea and approach to achieving CAPIS in UCE, so it maybe enrich theories of PIS, context-aware, extenics and social computing.

针对现有以少数专家为主导、预先定义的情境感知个性化方法不能满足泛在计算环境中随时随地个性化信息服务过程中海量多样化、动态变化的情境信息定义、提供、管理和利用问题，提出采用社会化驱动方式实现泛在计算环境中情境感知个性化的思路，建立一种以大众用户为主导、动态开放的情境感知个性化方法。在对社会化驱动的情境感知个性化的基本原理、协同机制和过程体系等基本理论研究基础上，构建情境和信息的事物元模型、关联空间模型和复杂超网络模型，提出情境和信息的自分类、自识别和自聚类等动态优化方法以及关联拓展、网络发现与优度评价相结合的情境发现获取方法和集成优化过程的匹配发现与优度评价相结合的信息发现获取方法，并构建系统进行评估应用。本项目研究基于应用需求，为泛在计算环境中情境感知个性化实现提供了新的思路与方法，丰富和完善情境感知个性化理论与方法，促进个性化信息服务、情境感知、可拓学和社会化计算等学科交叉形成自身特色。

随着泛在网络、传感技术和移动通信技术等的发展，一个以用户为中心的信息、物理和人类活动三者无缝融合的泛在计算环境正在形成，为用户提供无所不在的“按需”服务成为其关键问题。针对现有以少数专家为主导、预先定义的情境感知个性化方法不能满足泛在计算环境中随时随地个性化信息服务过程中海量多样化、动态变化的情境信息定义、提供、管理和利用问题，引入Web 2.0的用户参与理念和社会化标注方法，提出采用社会化驱动方式实现泛在计算环境中情境感知个性化的思路，以可拓学和社会化计算为主要理论和方法，建立一种以大众用户为主导、动态开放的情境感知个性化方法。项目研究较系统地建立了适应泛在计算环境的社会化驱动的情境感知个性化信息服务的基本理论，包括阐明了社会化驱动的情境感知个性化的基本原理，建立了社会化驱动的情境感知个性化的协同机制，构建了社会化驱动的情境感知个性化的过程与体系。提出并实现了一套适应泛在计算环境的社会化驱动的情境感知个性化的模型与方法，包括从节点、关联和作用关系三个层面建立了情境和信息资源的事物元模型、关联空间模型和复杂超网络模型并刻画了其关键特性；建立实现了情境和信息资源的自分类、自识别和自聚类等动态优化方法，用户情境信息发现获取方法以及情境感知的个性化信息资源发现获取方法等。构建适应泛在计算环境的社会化驱动的情境感知个性化信息服务系统，结合基准比较评价和实际应用案例展开评估应用研究，检验本项目提出的理论、模型和方法的可行性、正确性和有效性。本项目研究基于应用需求，为泛在计算环境中情境感知个性化实现提供了新的思路与方法，丰富和完善情境感知个性化理论与方法，促进个性化信息服务、情境感知、可拓学和社会化计算等学科交叉形成自身特色。