Deep hybrid collaborative filtering for Web service recommendation
Deep hybrid collaborative filtering for Web service recommendation
复制标题
用于Web服务推荐的深度混合协同过滤
DOI：
10.1016/j.eswa.2018.05.039
复制
发表时间：
2018-11-15
影响因子：
8.5
通讯作者：
Ma, Yutao
中科院分区：
文献类型：
--
作者：
Xiong, Ruibin;Wang, Jian;Ma, Yutao
With the rapid development of service-oriented computing and cloud computing, an increasing number of Web services have been published on the Internet, which makes it difficult to select relevant Web services manually to satisfy complex user requirements. Many machine learning methods, especially matrix factorization based collaborative filtering models, have been widely employed in Web service recommendation. However, as a linear model of latent factors, matrix factorization is challenging to capture complex interactions between Web applications (or mashups) and their component services within an extremely sparse interaction matrix, which will result in poor service recommendation performance. Towards this problem, in this paper, we propose a novel deep learning based hybrid approach for Web service recommendation by combining collaborative filtering and textual content. The invocation interactions between mashups and services as well as their functionalities are seamlessly integrated into a deep neural network, which can be used to characterize the complex relations between mashups and services. Experiments conducted on a real-world Web service dataset demonstrate that our approach can achieve better recommendation performance than several state-of-the-art methods, which indicates the effectiveness of our proposed approach in service recommendation. (C) 2018 Elsevier Ltd. All rights reserved.
