摘要
已有的智能信息推荐模型忽略了对兴趣特征向量的提取,导致推荐结果与用户要求不一致。MOOC平台包含海量的慕课课程信息,其信息推荐难度较大。为此,构建基于MOOC的智能信息推荐模型。针对MOOC平台用户学习需求,为用户推荐所需的慕课课程信息。根据层次差、语义距离以及语义重合度计算MOOC平台中信息本体的概念相似度。利用本体的概念相似度计算方法聚类MOOC平台中海量课程信息,通过深度神经网络方法通过卷积层、池化层以及输出层三个步骤,提取兴趣点特征向量,构建协同过滤图模型,实现MOOC平台智能信息推荐。仿真测试结果表明,该模型可利用所提取的兴趣点特征向量为用户推荐所需课程,课程推荐的平均排序倒数均为0.7以上,可实现MOOC平台的智能信息推荐。
At present,some intelligent information recommendation models ignore extracting interest feature vectors,leading to inconsistent recommendation results with user requirements.Massive Open Online Courses(MOOC)platform includes a large amount of MOOC information,so it is difficult to recommend information to users.Therefore,an intelligent information recommendation model based on MOOC was constructed.According to users' learning needs on MOOC platform,we can recommend MOOC information required by users.The conceptual similarity of information ontology on MOOC platform was calculated via hierarchical difference,semantic distance and semantic coincidence degree.In addition,we used the concept similarity calculation of ontology to cluster massive course information on MOOC platform.Based on convolution layer,pooling layer and output layer,we used deep neural network to extract the feature vectors of interest points and constructed a model of collaborative filtering graph.Finally,the intelligent information recommendation of MOOC platform was achieved.Simulation results show that the designed model can use the extracted feature vectors of interest points to recommend the courses required by users,and the mean reciprocal rank of course recommendation is always above 0.7.
作者
马莲姑
黄寿孟
纪春林
赵安学
MA Lian-gu;HUANG Shou-meng;JI Chun-lin;ZHAO An-xue(Qiongtai Normal University,School of Information Science and Technology,Haikou Hainan 571100,China;University of Sanya,School of Information and Intelligent Engineering,Sanya Hainan 572022,China;Shaanxi University of Technology,College of Educational Sciences,Hanzhong Shaanxi 723000,China)
出处
《计算机仿真》
北大核心
2023年第8期275-278,466,共5页
Computer Simulation
基金
海南省教育科学规划2021年度重点项目(QJZ20211007)
2022年海南省哲学社会科学规划课题(HNSK(ZC)22-156)
琼台师范学院科学研究项目(qtnb202103)。
关键词
智能信息
推荐模型
卷积层
兴趣点
语义距离
Intelligent information
Recommendation model
Convolutional layer
Interest points
Semantic dis-tance
