摘要
针对传统音乐推荐系统推荐精度不高、用户感知体验较差等问题,此次研究利用协同过滤算法和案例推理算法搭建了一种虚拟音乐推荐模型,并引入情境信息感知技术对该模型进行优化,改善其推荐效果。最后,搭配情境感知与用户交互模块,设计了完整的多模式虚拟音乐推荐系统。对算法性能和推荐系统应用效果进行测试,所设计算法的最佳召回率、精度值、F1值分别为0.981、0.978、0.985。整个推荐系统更是得到了0.92和0.83的用户满意度与专家满意度。综上,此次研究所设计的虚拟音乐推荐系统拥有较好的推荐效果,能够用于实际应用中。
To solve the problems of low recommendation accuracy and poor user perception experience of traditional music recommendation systems,this research uses collaborative filtering algorithm and case-based reasoning algorithm to build a virtual music recommendation model,and introduces situational information awareness technology to optimize the model to improve its recommendation effect.Finally,a complete multimodal virtual music recommendation system was designed in conjunction with situational awareness and user interaction modules.The performance of the algorithm and the application effect of the recommendation system were tested,and the optimal recall rate,accuracy value,and F1 value of the designed algorithm were 0.981,0.978,and 0.985,respectively.The entire recommendation system achieved user satisfaction of 0.92 and expert satisfaction of 0.83.In summary,the virtual music recommendation system designed by this research institute has good recommendation performance and can be used in practical applications.
作者
党焱
DANG Yan(Shaanxi Polytechnic Institute,Xianyang,Shaanxi 712000,China)
出处
《自动化与仪器仪表》
2024年第1期30-34,共5页
Automation & Instrumentation
基金
陕西工业职业技术学院2022年校级科研项目《“大思政”格局下高职院校美育与思政教育有机融合的探索与实践》(2022YKYB-068)。
关键词
虚拟音乐
案例推理
感知
交互设计
情境
协同过滤
virtual music
case based reasoning
perception
interaction design
context
collaborative filtering
