基于服务关联模型的服务排序算法——ServiceRank

朱怡安; 雷万保

doi:10.3969/j.issn.1001-0548.2011.04.027

基于服务关联模型的服务排序算法——ServiceRank

朱怡安¹,
雷万保²

1.
西北工业大学软件与微电子学院西安 710072;
2.
西北工业大学计算机学院西安 710072

基金项目:

部级基础科研项目;部级预研基金

详细信息

作者简介:
朱怡安(1961-)，男，教授，主要从事软件工程、计算机网络和服务计算方面的研究.

中图分类号: TP306.1
计量
- 文章访问数: 4549
- HTML全文浏览量: 156
- PDF下载量: 73
出版历程
- 收稿日期: 2009-11-12
- 修回日期: 2010-12-17
- 刊出日期: 2011-08-14

Service Ranking Algorithm Based on Service's Relational Model—ServiceRank

ZHU Yi-an¹,
LEI Wan-bao²

1.
School of Software and Micro-electronics,Northwest Polytechnical University Xi'an 710072;
2.
School of Computer Science and Technology,Northwestern Polytechnical University Xi'an 710072

摘要

摘要: 提出一种以Web service关联结构为基础的权重计算和排序算法——ServiceRank。在ServiceRank中，每个Webservice借助后继可组合关联关系，将自身的权重以连接度为尺度平均地分配给所有后继可组合Web service，并以类似的方法从所有前驱可组合Web service中获取权重。与现有方法相比，ServiceRank能提供全局一致的权重函数，且能离线工作，算法稳定性好且效率高。通过实验对算法的收敛性和稳定性进行了深入分析。
- 服务组合 /
- 服务排序 /
- 服务关联结构 /
- 面向服务的计算 /
- Web服务
Abstract: On the basis of Web service relational structure, this paper proposes an algorithm called ServiceRank to compute the service weight and rank the services In ServiceRank, by analyzing the relationship among the subsequent composable Web services, each Web service averagely distributes its weight by connectivity to all the subsequent composable Web services, and captures weight from all the predecessor composable Web services through the similar method. Compared with the current methods, ServiceRank has the ability to compute the unified weight function and to work offline. This paper thoroughly analyzes the convergence and stability of SercviceRank through experiments, which proves that the proposed algorithm has a high stability and efficiency.
- service composition /
- service ranking /
- service relational structure /
- SOC /
- Web service