【作者】 杨巍；

【导师】 陈莘萌；

【作者基本信息】 武汉大学 ， 计算机软件与理论， 2010， 博士

【摘要】 随着信息技术与社会科学的飞速发展,适应性学习成为电子学习发展的新趋势。在进行适应性学习时,学习者把已有的知识结构作为认知基础,通过与适应性学习系统进行交互的方式获取新知识,从而完成知识结构的强化或更新。适应性学习充分考虑学习行为的个性化特征,为学习者提供个性化学习策略与个性化学习内容,促进学习者知识的自我建构。近年来,关于适应性学习的研究逐渐深入,支持适应性学习的工具与系统陆续出现,系统的适应性与有效性不断提高,但是在研究和应用的发展过程中,仍存在以下一些问题与困难。1)缺少能对适应性学习策略进行形式化描述的方法与模型,没有学习目标的指引,容易出现学习迷航现象。2)学习者特征模型的准确性与有效性有待增强,在描述学习者个性化特征时的人为主观性与角度多样性决定了在构建学习者特征模型时选择与量化其特征项的困难性。3)学习内容获取的智能性需要进一步提高,学习效果取决于获得的学习内容是否能够适应学习者特征和学习需求。本文深入研究了适应性学习系统的相关问题,主要包括适应性学习设计的形式化描述、在建构主义理论指导下学习者知识刻画模型的建构、基于本体的学习资源语义描述与适应性推荐以及适应性学习系统的结构与功能。本文的研究工作主要有以下五项创新点。1)使用IMS学习设计对适应性学习的学习策略进行形式化描述,并引入领域概念组件对IMS学习设计描述知识结构的能力进行扩充与改进,让学习设计表达的信息更加全面、直观,便于学习者在学习过程中准确把握学习目标,并为更准确推断学习者的学习需求打下基础。2)对学习者特征项中的认知状态进行深入研究,提出了一种学习者知识刻画模型及知识点学习状况计算方法,提高了描述学习者特征的准确性。3)提出了一种基于学习者知识刻画模型的学习路径提取算法,并归纳了学习路径上知识流转趋势的类型,提出了一种通过分析学习路径上知识流转趋势生成学习方案的算法,提高了获取学习者学习需求的能力。4)提出了一种基于语义相似度的学习内容推荐算法,可实现学习内容的自动推荐。5)将学习设计和本体应用于适应性学习系统,提出了一种以适应性学习设计为目标导向、以本体为推理基础的适应性学习系统模型与运行机制。更多还原

【Abstract】 With the rapid development of information technology and social science, adaptive learning becomes the main trend of e-learning. During adaptive learning, learners obtain knowledge and ability through themselves’ knowledge experience and adaptive learning system. Adaptive learning thinks over the individual character sufficiently, so it can provide the individual learning strategy and learning content, promote learners’ knowledge self-constructing. Recently, researches about adaptive learning become profounder; tools and systems supported adaptive learning have appeared in succession. All of these can raise the adaptability and-validity of the system. But there are still some problems and difficulties during the research process. The main items are as follows.1) Lacking the method and model about describing of adaptive learning’s strategy formally. Learners can easily get lost in learning without learning strategy.2) The learner’s character model should be more precise and effectual. The diversity and changeability of learners’individual character items determine the difficulty of constructing learners’character model.3) The intelligence of learning content obtaining should be raised. The learning effect lies on whether the learning content obtained suits the learners’characters and learning requirement or not.This paper researches the relative problems above profoundly. The main contents are: the formal description about adaptive learning design; constructing learners’knowledge profiling model under the direction of constructivism; learning resource’s semantic describing and adaptability recommending based on ontology; structure and function of adaptive learning system.There are five innovative points in this research work.1) Using IMS-LD to describe the adaptive learning strategy formularly, and bringing in knowledge components to extend and improve the IMS-LD, so that make the information expressed by learning design more complete and intuitionistic, and direct learners to avoid getting lost during learning because that the learning target is not clear, then make the base for deducing the learners’ learning needs more correctly. 2) Making deeper research on the cognitive status in the learning character items, putting forwards the learners’knowledge profiling model and computing methods of knowledge learning status, all of these can improve the precision of describing learners’ characters.3) Designing and realizing the extracting path algorithm during learners knowledge profiling; Summing up the types of knowledge transfer trends; Designing and realizing the algorithm about deducing learners’ needs through analyzing knowledge transfer trends, so that can improve the capability of obtaining learners’learning needs.4) Designing and realizing the algorithm of learning content recommending based on semantic comparability, combining with the semantic description about learning resource, so that can increase the validity of learning content recommending.5) Apply learning design and ontology to the adaptive learning system; bringing forwards the adaptive learning system model and process mechanism on the basis of learning design and ontology.更多还原