摘要
针对学术交流资源数量与日俱增,学者自身研究方向多样性、学者与资源的交互行为稀疏等问题,提出考虑学科交叉需求的学术交流资源推荐方法。融合协同过滤与知识图谱算法,通过优化的TransE模型对学术交流资源知识进行表示学习,基于资源语义向量计算学术交流资源语义相似度,分析学者对学科交叉研究的需求,基于交互行为计算学者相似度,形成学者对学术交流资源的感兴趣程度,进而得出最终推荐结果。结果表明,相比传统推荐协同过滤推荐算法,该算法拥有较高的性能。
To solve the increasing number of academic exchange resources,the diversity of scholars′research directions,and the sparse interaction between scholars and resources,a method of academic communication resources recommendation based on the characteristics of interdisciplinary was proposed.Fusing collaborative filtering with knowledge graph algorithms,through the optimized TransE model,the knowledge of academic exchange resources was expressed and learned,and the semantic similarity of academic exchange resources was calculated based on the resource semantic vectors.At the same time,considering the needs of scholars for interdisciplinary research,the similarity of scholars was calculated based on the interaction behavior,and the degree of interest of scholars in academic exchange resources was further formed,so as to obtain the final recommendation results.Experimental results show that this algorithm has higher performance than the traditional recommendation collaborative filtering recommendation algorithm.
作者
栾玫琳
姜宁
李家普
魏勤
LUAN Meilin;JIANG Ning;LI Jiapu;WEI Qin(School of Information Engineering,Wuhan University of Technology,Wuhan 430070,China;不详)
出处
《武汉理工大学学报(信息与管理工程版)》
CAS
2024年第1期159-163,共5页
Journal of Wuhan University of Technology:Information & Management Engineering
基金
武汉理工大学自主创新研究基金项目(213275005)
科技部重点研发项目(20201f0092,202009YF01)。
关键词
学科交叉
学术交流资源
协同过滤
知识图谱
推荐算法
interdisciplinary
academic communication resources
collaborative filtering
knowledge graph
recommendation