随着互联网科技日新月异的发展,网络中存储的数据总量正以前所未有的速度激增。在这种背景下,如何从海量数据资源中精准高效地提炼出我们所需要的特定信息,已然成为当下亟待解决的重要议题。在这项研究中,我们提出了一种新的协同过滤算...随着互联网科技日新月异的发展,网络中存储的数据总量正以前所未有的速度激增。在这种背景下,如何从海量数据资源中精准高效地提炼出我们所需要的特定信息,已然成为当下亟待解决的重要议题。在这项研究中,我们提出了一种新的协同过滤算法来生成电子商务产品中的推荐。这项研究有两个主要创新之处。首先,我们提出了一种嵌入时间行为信息的机制,以找到一个邻居集,其中每个邻居对当前用户或项目有非常重要的影响。其次,在概率矩阵分解中引入邻居集,提出了一种新的协同过滤算法。我们将所提出的方法与实际数据集上的几种最先进的替代方法进行了比较。实验结果表明,本文提出的方法优于现有的方法。With the relentless and groundbreaking advancements in internet technology, the aggregate volume of data stored within network architectures is expanding at an unparalleled velocity. In this context, the task of precisely and efficiently extracting the targeted information from the deluge of data resources has emerged as a crucial and pressing issue demanding immediate attention and resolution in contemporary times. In this study, we propose a new collaborative filtering algorithm to generate recommendations in e-commerce products. There are two main innovations in this study. First, we propose a mechanism for embedding temporal behavior information to find a set of neighbors, where each neighbor has a very important influence on the current user or project. Secondly, a new collaborative filtering algorithm is proposed by introducing neighbor sets into probability matrix factorization. We compared the proposed method with several state-of- the-art alternatives on actual datasets. Experimental results show that the proposed method is superior to the existing methods.展开更多
文摘随着互联网科技日新月异的发展,网络中存储的数据总量正以前所未有的速度激增。在这种背景下,如何从海量数据资源中精准高效地提炼出我们所需要的特定信息,已然成为当下亟待解决的重要议题。在这项研究中,我们提出了一种新的协同过滤算法来生成电子商务产品中的推荐。这项研究有两个主要创新之处。首先,我们提出了一种嵌入时间行为信息的机制,以找到一个邻居集,其中每个邻居对当前用户或项目有非常重要的影响。其次,在概率矩阵分解中引入邻居集,提出了一种新的协同过滤算法。我们将所提出的方法与实际数据集上的几种最先进的替代方法进行了比较。实验结果表明,本文提出的方法优于现有的方法。With the relentless and groundbreaking advancements in internet technology, the aggregate volume of data stored within network architectures is expanding at an unparalleled velocity. In this context, the task of precisely and efficiently extracting the targeted information from the deluge of data resources has emerged as a crucial and pressing issue demanding immediate attention and resolution in contemporary times. In this study, we propose a new collaborative filtering algorithm to generate recommendations in e-commerce products. There are two main innovations in this study. First, we propose a mechanism for embedding temporal behavior information to find a set of neighbors, where each neighbor has a very important influence on the current user or project. Secondly, a new collaborative filtering algorithm is proposed by introducing neighbor sets into probability matrix factorization. We compared the proposed method with several state-of- the-art alternatives on actual datasets. Experimental results show that the proposed method is superior to the existing methods.
