【作者】 刘晓东；

【导师】 仲秋雁；

【作者基本信息】 大连理工大学 ， 技术经济及管理， 2013， 博士

【摘要】 普适计算技术在计算机技术和移动通信技术发展的双重推动下,已经渗透到现实生活中的各个领域,正在为人类创建了一个随时随地获取信息的数字空间,也使普适学习成为可能。普适学习区别于传统e-learning的主要特征就是移动性带来的情境的变化,普适学习系统必须能够感知这些变化,通过推理等过程做出判断,主动地、智能地调整系统行为,最终为学习者提供适应其情境的学习体验。普适学习研究的核心问题是学习情境,包括学习者情境、学习资源情境、位置情境等,因此普适学习系统的核心功能就是对情境数据的处理过程,即情境管理。本文基于普适学习情境的分析,设计情境模型,研究其获取、聚合、推理、分发以及质量控制等管理过程,构建情境管理架构,提出了情境数据质量评价指标体系及量化方法。主要的研究内容及成果包括：(1)普适学习情境分析。对情境的分类法进行总结和分析,提出了新的普适学习情境分类体系,并对学习者、学习资源、学习者位置三个情境进行了重点分析,研究了其获取方法。普适学习情境分类体系及获取方法研究,为情境的管理提供了基础。(2)基于本体情境建模方法的优势分析,采用本体方法构建普适学习情境模型,分别设计了普适学习系统的学习者、学习资源、学习者位置本体模型,并设计了普适学习情境总体模型和模型驱动方法。情境本体模型为情境管理提供了结构化的情境数据中心,并成为基于本体的情境推理的基础。(3)设计了面向终端学习设备的情境管理架构,分别对相关层次和模块进行功能设计,包括外围模块(情境采集、情境提供者和消费者层次)和情境管理核心模块(包括情境聚合、抽象和推理、质量控制以及分发层次),这一架构集成了本体推理和基于规则的推理,综合了本体推理的功能优势和规则推理的效率优势,同时将情境质量控制贯穿到了情境管理的全过程。(4)针对情境管理研究中的一个相对薄弱环节,对情境数据的质量评价做了专门研究,使用德尔菲法构建了一套情境质量评价指标体系,并分别设计了量化方案,情境质量评价指标体系为情境质量控制提供了保障。更多还原

【Abstract】 With the promotion of computer technique and mobile communication technology, pervasive computing technology has penetrated in all the fields of real life and is creating a digital space in which human beings can get access to information whenever and wherever possible and then makes pervasive learning possible. The main characteristic of pervasive learning which is different from traditional e-learning is the changes of context which is brought from mobility. Pervsive learning system must perceive these changes, do the right choices through reasoning and adjust system behaviours actively and intelligently and finally appropriate learning experiences for learners.The main core questions of pervasive learning study is learning context, including learner context, learning resource context, location context and so on. So the core function of pervasive learning system is the processing of context data, that is context management. Based on analysis on pervasive learning context, this paper designs context model and studies its acquisition, aggregation, reasoning, dissemination and quality control and some other management processes. Then this paper tends to create context management architecture and bring forward context quality evaluating system and quantification method.The main contents and products include:(1) Analysis on pervasive learning context.It concludes and analyzes the classification of contexts, brings forward a new pervasive learning context classification system and does a specific study on learner context,learning resources and learner location, finally studies its acquisition method. Study on pervasive learning context classification system and acquisition method promote a foundation for context management.(2) On the basis of advantage analysis of ontology context modeling method, it applies ontology method to create pervasive learning context model and designs ontology models of learners,learning resources and learner location of pervasive learning system and also designs general pervasive learning model and model generating method. Context ontology model promotes a constructive data center for context regulating and becomes a foundation for context reasoning basen on ontology.(3) It designs a terminal learning device-oriented context management architecture, and designs the functions for corresponding classes and units, which include peripheral unit(context collecting, context provider and consumer layer) and core context management unit(context aggregation, abstracting and reasoning, quality control and dissemination layer). This architecture integrates ontology reasoning and reasoning based on rules and combines the function advantages of ontology reasoning and efficiency advantage of rule reasoning and it also put context quality control into all the process of context management.(4) As to a relatively weak process of context management study, it does a specific study on the quality evaluation on context data and brings a set of context quality evaluating system and designs quantification plans individually. Quality of Context evaluating indicator system promote security for context quality control.更多还原