知识服务云平台下个性化学习机制研究

发布时间：2018-04-09 10:08

本文选题：知识服务　切入点：云计算　出处：《南京邮电大学》2013年硕士论文

【摘要】：社会分工更加细化，人们需要具备更加专业化和精准化的知识，而传统的信息服务基点、重点和终点则是信息资源的获取，无法提供真正意义上的按需服务、个性化服务；知识更新速度加快，但学习者在网络学习的过程中，存在学习迷航和信息超载两大问题；云计算技术逐步成熟，推动各领域迅速发展，将云计算技术与知识服务深入的结合在一起，能够为学习者提供更好的服务。通过剖析知识服务的内涵和特征，梳理知识服务中存在的问题，结合云计算的特性，给出共享学习模式下知识服务的新模式。同时指出基于云计算的知识服务需要解决的三个关键技术问题，即知识服务云平台的构建、个性化的学习路径推荐和知识服务计费。首先构建了知识服务云平台体系结构，设计了知识资源层、知识融合层、知识核心服务层、运营服务与计费保障层和客户应用层五个知识服务功能模块，实现知识资源的共享和协同。其次为了能够借助云计算平台为学习者更好的提供知识服务，提出了基于蚁群优化算法的个性化学习路径推荐策略。该策略中，将与目标用户具有相似学习风格和知识水平的历史学习者的评价信息作为信息素，学习对象的表达特征和难易度作为启发信息来引导路径搜索行为，最终将推荐概率较高的学习路径推送给学习者。通过模拟实验调整相关参数提升推荐精度，证明了算法的有效性和可行性。最后通过分析云计算环境下服务模块流程组合的动态性，，运用随机Petri网进行动态流程建模与追踪，给出了分层计费体系下的流程计费模型，使用Petri网对该计费流程进行建模分析和设计，便于云服务提供商实现基于状态的计费控制，从而优化计费策略。实现云计算基于过程模型的服务计量，根据该模型进行相应定价，用户即可按需付费，通过应用实例验证了该计费方法的有效性和可操作性，为云服务提供商计费决策提供支撑。
[Abstract]:Social division of labor is more detailed, people need to have more specialized and accurate knowledge, and traditional information service base point, the focus and end point is the acquisition of information resources, can not provide a true sense of on-demand services, personalized services;The speed of knowledge updating is quickening, but in the process of online learning, there are two major problems of learning confusion and information overload. Cloud computing technology is gradually maturing and promoting the rapid development of various fields.The combination of cloud computing technology and knowledge service can provide better service for learners.By analyzing the connotation and characteristics of knowledge service, combing the existing problems in knowledge service and combining the characteristics of cloud computing, a new model of knowledge service in shared learning mode is proposed.At the same time, it points out three key technical problems that need to be solved in knowledge services based on cloud computing, namely, the construction of knowledge service cloud platform, personalized learning path recommendation and knowledge service accounting.Firstly, the architecture of knowledge service cloud platform is constructed, and five knowledge service function modules are designed, including knowledge resource layer, knowledge fusion layer, knowledge core service layer, operation service and billing support layer and customer application layer.To realize the sharing and cooperation of knowledge resources.Secondly, in order to provide better knowledge service for learners with the help of cloud computing platform, a personalized learning path recommendation strategy based on ant colony optimization algorithm is proposed.In this strategy, the evaluation information of historical learners with similar learning style and knowledge level is used as pheromone, and the expression characteristics and ease of learning objects are used as heuristic information to guide the path search behavior.Finally, the recommended learning path with high probability is pushed to the learner.The effectiveness and feasibility of the algorithm are proved by adjusting the relevant parameters to improve the recommendation accuracy through simulation experiments.Finally, by analyzing the dynamic characteristics of service module process composition in cloud computing environment, using stochastic Petri net to model and track the dynamic process, a process billing model based on hierarchical billing system is presented.The Petri net is used to model and design the charging process, which is convenient for cloud service providers to realize state-based billing control and optimize billing strategy.The service measurement based on the process model is realized, according to the model, the user can pay according to the model. The validity and maneuverability of the charging method are verified by an application example.Provides the support for the cloud service provider billing decision.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F49

【参考文献】