摘要
为了解决传统协同过滤算法在捕捉用户兴趣动态变化方面的不足,本研究提出了一种基于艾宾浩斯遗忘规律的动态协同过滤算法。该算法通过引入时间衰减函数后,能够根据用户评价的时间点动态调整评价权重,并成功地将静态评价矩阵转化为动态评价矩阵。实验结果表明,本算法实现了平均精确率的显著提升,最高可达1,而传统协同过滤算法的平均精确率通常在0.1至0.6之间,这充分表明了改进算法在推荐准确性方面的优越性。因此,本算法具有广泛的应用前景,有望为电子商务、在线媒体平台等多个领域带来更加精准和个性化的推荐服务。
To address the limitations of traditional collaborative filtering algorithms in capturing the dynamic changes in user interests,this study proposes a dynamic collaborative filtering algorithm based on the Ebbinghaus Forgetting Curve.Specifically,by introducing a time-decay function,the algorithm can dynamically adjust the evaluation weights according to the time points of user ratings,successfully converting the static evaluation matrix into a dynamic one.Experimental results show that this algorithm significantly enhances the recommendation accuracy.The average precision of traditional collaborative filtering algorithms usually fluctuates between 0.1 and 0.6.In contrast,our algorithm significantly boosts the average precision,reaching up to 1.This clearly shows the superior performance of the improved algorithm in terms of recommendation accuracy.Therefore,this algorithm holds broad application prospects and is expected to bring more precise and personalized recommendation services to various fields such as e-commerce and online media platforms.
作者
王方圆
郝奎
张国华
WANG Fangyuan;HAO Kui;ZHANG Guohua(School of Science,Hunan University of Technology,Zhuzhou,Hunan 412000,China)
出处
《湖南城市学院学报(自然科学版)》
2025年第2期74-78,共5页
Journal of Hunan City University:Natural Science
关键词
艾宾浩斯遗忘规律
协同过滤
时间因素
推荐算法
准确度
Ebbinghaus Forgetting Law
collaborative filtering
time factor
recommendation algorithms
accuracy
