Generating personalized study guide for self-paced learning online

在线课程的学生通常遵循预先计划的学习路径，通常由老师以学习指南的形式规定。学生在阅读、教材、教学视频剪辑和动画、练习、作业、自我评估和反思方面遵循学习指南，通过调查、讨论、团队工作和项目与同龄人合作，通过模拟试验、测验成绩和游戏表演与他人竞争。然而，学习指南并不能满足所有学生的需求。每个学生都有自己的学习偏好、节奏和轨迹。一个单一的，固定的，预先设计的学习指南也不能为所有学生提供有效和高效学习的最佳方法。

Students in online courses typically follow a pre-planned learning path, typically prescribed by a teacher in the form of a study guide. Students follow the study guide for readings, teaching materials, instructional video clips and animations, practice exercises, assignments, self-assessments and reflections, collaborating with peers via survey, discussion, team works and projects, and competing with others via simulation trials, quiz scores and game-play performances. However, a study guide cannot meet all students' needs. Every student has individual learning preferences, paces, and trajectories. Nor can a single, fixed, pre-designed study guide provide all students with the best way to learn effectively and efficiently. This research aims to enable an open-source learning management system (i.e., Moodle and edX) to provide students with a personalized study guide for their online learning via the use of graph structures to store and represent the learning objects in an online course as well as the learning behaviours, strategies, and preferences of individual students studying online. The research first finds the relations among the learning objects in an entry level, online Physics course. This research then designs a learning analytics approach to process students' behaviours that are recorded by the learning management system and automatically categorize students into clusters according to the degrees of similarity of students' learning strategies and preferences while learning online. This research is expected to result in automatic re-planning/re-arrangement of not only learning objects (i.e., video clips, hands-on, simulations, mini-games, real-life problems, etc.), but also the learning trajectory (i.e., sequence of learning concepts, topics, and units) for the individual student according to the analysis result of the student's learning behaviours. More specifically the outcome of this research can be summarized as: (1) a graph structure that can be used to store and present the relations among learning objects in an online course; (2) a learning analytics approach that analyzes students' time-series learning behaviours in a learning management system/platform, finds students' behaviour patterns based on their learning strategies and preferences; (3) a method that generates a personalized learning trajectory for the individual student according to their student model and behaviour pattern; and, (4) personalized study guide plug-ins for open-source learning management systems and MOOC platforms (e.g., Moodle and edX) so that any university, college, or facility offering online and/or MOOC courses can adopt the plug-in into its system/platform and students can benefit from the results of this research.