基于蚁群信息素优化算法的微学习路径推荐研究

发布时间：2018-01-06 13:39

本文关键词：基于蚁群信息素优化算法的微学习路径推荐研究　出处：《太原理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：微学习是一种新型学习方式,是在线学习适应碎片化时代的一种发展形式。微学习最主要的特点是,其学习单元包含的知识内容相对精简,构成包括文本、音频/视频和图像等多种形式,使学习过程中学习者在时间和空间上受到的限制较少。短小精悍的学习单元,灵活多变的组织方式,自由开放的学习环境使得微学习一经提出,便受到了广泛关注。与其他在线学习方式一样,由于学习资源的激增,使微学习也面临着“信息过载”问题。学习者不得不在寻找适合的学习单元时花费大量的时间,从而影响学习效率。以上问题的存在,促使我们寻找一种方法来为学习者提供适合的微学习单元。本文在充分调查研究在线学习相关技术的基础上,结合微学习的特征,提出了一种蚁群信息素优化算法来推荐微学习单元,并通过逐步适应的方式实现对微学习路径的推荐。该方法将微学习特征与蚁群算法有机结合,以有效提高学习者在微学习中的学习效率。在介绍学习路径推荐的相关概念,和分析微学习概念与特性的基础上,本文对学习单元属性、学习者特征和学习路径进行分析建模,并对其表示方式进行定义,进而提出了微学习中学习路径推荐模型。本文提出的微学习路径推荐方法的核心是蚁群信息素优化算法,结合微学习路径推荐方法的整体设计框架,以及对框架中的各个模块的功能描述,详细介绍了该算法实时监测学习者的学习情况变化,并使用信息素浓度系数自适应函数调节信息素浓度,及时优化调整微学习路径推荐策略的处理流程。其后,通过MATLAB环境,得到适用于微学习环境的蚁群算法参数组合。最后通过实验验证了该方法可以为微学习中学习者推荐适合的学习路径,有效提高学习效率。在充分分析微学习和在线学习共同面临问题的基础上,本文结合微学习的特点,提出一种基于蚁群信息素优化算法的微学习路径推荐方法。该方法的主要创新点如下:1.分析微学习的特点,借助蚁群算法反馈性的优势,将学习路径推荐粒度细化为更小的学习单元,进一步提高了推荐的精度。2.分析学习过程和学习要素,借助蚁群算法中的蚂蚁属性分类,监测学习过程,根据学习者学习状态的变化,逐步调整推荐策略,优化学习路径,以适应学习者的需求变化。3.深入分析学习流程,引入信息素浓度系数自适应调节函数,对全局和局部信息素浓度进行调节,优化推荐结果。
[Abstract]:Micro-learning is a new learning method, which is a development form of online learning to adapt to fragmentation. The main feature of micro-learning is that its learning unit contains relatively simple knowledge content, including text. Audio / video and image and other forms, so that learners in the learning process in the space and time constraints are less, short learning units, flexible organization. The free and open learning environment makes micro-learning, once proposed, has attracted wide attention. Like other online learning methods, because of the proliferation of learning resources. Micro-learning is also facing the problem of "information overload". Learners have to spend a lot of time looking for suitable learning units, thus affecting learning efficiency. This paper makes us find a way to provide suitable micro-learning units for learners. This paper combines the characteristics of micro-learning on the basis of full investigation and research of online learning related technologies. In this paper, an ant colony pheromone optimization algorithm is proposed to recommend the micro-learning unit, and the micro-learning path is recommended by gradual adaptation, which combines the micro-learning features with the Ant Colony algorithm. In order to effectively improve the learning efficiency of learners in micro-learning. On the basis of introducing the related concepts of learning path recommendation and analyzing the concept and characteristics of micro-learning, this paper deals with the attributes of learning units. Learner characteristics and learning paths are analyzed and modeled, and their representation is defined. Then the learning path recommendation model in micro-learning is proposed. The core of the proposed micro-learning path recommendation method is the ant colony pheromone optimization algorithm combined with the overall design framework of micro-learning path recommendation method. As well as the functional description of each module in the framework, the algorithm is introduced in detail to monitor learners' learning changes in real time, and the pheromone concentration coefficient adaptive function is used to adjust the pheromone concentration. Optimize and adjust the process of micro-learning path recommendation strategy in time. Then, through the MATLAB environment. An ant colony algorithm parameter combination suitable for micro-learning environment is obtained. Finally, the experimental results show that the proposed method can recommend suitable learning paths for micro-learning learners. Effectively improve learning efficiency. On the basis of fully analyzing the common problems of micro-learning and online learning, this paper combines the characteristics of micro-learning. A microlearning path recommendation method based on ant colony pheromone optimization algorithm is proposed. The main innovation of this method is as follows: 1. The characteristics of microlearning are analyzed and the advantage of feedback is obtained. The recommended granularity of learning path is refined into a smaller learning unit, which further improves the accuracy of recommendation. 2. Analyze the learning process and learning elements, and monitor the learning process by means of ant attribute classification in ant colony algorithm. According to the change of learner's learning state, the recommendation strategy is adjusted step by step, and the learning path is optimized to adapt to the change of learner's demand. 3. Deeply analyze the learning process and introduce the adaptive adjustment function of pheromone concentration coefficient. The global and local pheromone concentrations were adjusted to optimize the recommended results.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18

【参考文献】