Please use this identifier to cite or link to this item:
Scopus Web of Science® Altmetric
Type: Journal article
Title: Unified collaborative and content-based web service recommendation
Author: Yao, L.
Sheng, Q.Z.
Ngu, A.H.H.
Yu, J.
Segev, A.
Citation: IEEE Transactions on Services Computing, 2015; 8(3):453-466
Publisher: IEEE
Issue Date: 2015
ISSN: 1939-1374
Statement of
Lina Yao, Quan Z. Sheng, Anne. H.H. Ngu, Jian Yu, and Aviv Segev
Abstract: The last decade has witnessed a tremendous growth of web services as a major technology for sharing data, computing resources, and programs on the web. With increasing adoption and presence of web services, designing novel approaches for efficient and effective web service recommendation has become of paramount importance. Most existing web service discovery and recommendation approaches focus on either perishing UDDI registries, or keyword-dominant web service search engines, which possess many limitations such as poor recommendation performance and heavy dependence on correct and complex queries from users. It would be desirable for a system to recommend web services that align with users’ interests without requiring the users to explicitly specify queries. Recent research efforts on web service recommendation center on two prominent approaches: collaborative filtering and content-based recommendation. Unfortunately, both approaches have some drawbacks, which restrict their applicability in web service recommendation. In this paper, we propose a novel approach that unifies collaborative filtering and content-based recommendations. In particular, our approach considers simultaneously both rating data (e.g., QoS) and semantic content data (e.g., functionalities) of web services using a probabilistic generative model. In our model, unobservable user preferences are represented by introducing a set of latent variables, which can be statistically estimated. To verify the proposed approach, we conduct experiments using 3,693 real-world web services. The experimental results show that our approach outperforms the state-of-the-art methods on recommendation performance.
Rights: © 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
RMID: 0030031138
DOI: 10.1109/TSC.2014.2355842
Appears in Collections:Computer Science publications

Files in This Item:
File Description SizeFormat 
RA_hdl_95084.pdfRestricted Access669.57 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.