基于知识图谱及案例推理的MOOC资源推荐算法研究  

Research on the Algorithm of MOOC Recommendation Based on Knowledge Map and Case-based Reasoning

作　　者：门瑞[1] 杨芳萍[1] MEN Rui;YANG Fang-ping(College of Information and Engineering, Longdong University, Qingyang 745000, Gans)

机构地区：[1]陇东学院信息工程学院,甘肃庆阳745000

出　　处：《陇东学院学报》2018年第1期10-13,共4页Journal of Longdong University

基　　金：甘肃省教育科学‘十三五’规划2016年度<基于云计算的甘肃省高校教学资源共享及推荐平台研究(GS[2016]GHBZ021)

摘　　要：在对各种推荐算法分析研究的基础上,经过多次试验,得出适合MOOC资源的推荐算法,实现了较好的推荐效果。对于使用过MOOC资源的用户,根据其历史记录,查找已使用资源在知识图谱中的位置,推荐满足个性化需求的后续资源;对于未使用过MOOC资源的用户,根据其选定的知识领域,利用基于案例推理的推荐算法,在案例库中查找相同或相似度最高的案例,根据案例中用户的选择,为其推荐最优的资源。此算法解决了国内MOOC平台中存在的没有推荐模块或推荐结果不合理的问题,在某实验平台上取得了良好的推荐效果。Based on the analyses research of several recommendation algorithms and experiments, this paper finds out an algorithm which works effectively on recommendation of MOOC resources. In detail, for those who used to access MOOC resources, the algorithm checks their location on the knowledge map according to their history records and recommends them resources which meet their personalized requirements. For those who do not have access history, according to the knowledge area they choose, the algorithm firstly applies a case-based reasoning to find the most analogical case and then based on the choices of the users in the case, recommends them the best resources. Generally, the algorithm solves the problem that the most MOOC platforms do not have recommendation module or unreasonable module, and gets good recommendation effects on an experimental platform.

关 键 词：知识图谱 案例推理 慕课资源 推荐算法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]