• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2014, Vol. 36 ›› Issue (04): 713-718.

• 论文 • 上一篇    下一篇

基于NEI的免疫自学习Web服务突现方法研究

郭兰兰,王辉,陈翔涛   

  1. (河南科技大学电子信息工程学院,河南 洛阳 471023)
  • 收稿日期:2012-08-29 修回日期:2013-01-15 出版日期:2014-04-25 发布日期:2014-04-25
  • 基金资助:

    河南省教育厅自然科学研究项目(2010A520017);河南省教育厅科学技术研究重点项目(2012B520019)

Research of immune self-learning Web service emergence method
based on biological neuro-endocrine immune system        

GUO Lanlan,WANG Hui,CHEN Xiangtao   

  1. (School of Electronic Information Engineering, Henan University of Science and Technology,Luoyang 471023,China)
  • Received:2012-08-29 Revised:2013-01-15 Online:2014-04-25 Published:2014-04-25

摘要:

针对复杂任务的Web服务组合问题,借鉴生物神经内分泌免疫(NEI)系统的突现和自学习机制,提出了一种免疫自学习服务突现方法。移动Agent设计为具有免疫行为的生物实体,并代理Web服务。突现的服务组合是生物实体Agent通过亲合力匹配算法形成的突现实体网络提供的,并且能够动态地调整其内部的服务。采用免疫原理和自学习机制,将亲和力匹配形成的服务组合视为抗体进行记忆。当再次遇到相同或相似服务请求时,将直接进行二次应答或对抗体库中的抗体进行修正组合,形成新的中间抗体,从而更加快速且节省能量地完成服务响应。实验表明,该方法能够自组织地完成服务的动态组合、自主学习和管理等工作,而且可以提高响应速度和服务效率。

关键词: Web服务组合, 语义Web服务, 突现, 亲和力匹配, 免疫学习

Abstract:

In order to resolve the combination problems of Web service for complicated tasks, inspired by the characteristics of system emergence and selflearning in biological neuroendocrine and immune system, a method of immune learning Web service emergence is proposed. Mobile agent is designed as a biological entity with immune behaviors and acts for Web services. Emergent bioentity network is constructed through bioentities affinity matching and provides emergent services composition, in the process, service can be added or removed dynamically in service emergence. Based on immune principle and selflearning mechanism, affinity matching service combination is viewed as antibody for memory. When system encounters the same or similar service requests, it directly forms the new secondary response or updates antibody library to form the new intermediate antibody, and completes the response in a more rapid and energyserving way. Through experiment verification, the method can complete selforganizing dynamic combination, independent learning and management, and improve the response speed and efficiency of services.

Key words: web services composition;semantic web service;emergence;affinity matching;immune learning