一个面向Internet的分布式信息检索系统模型

A Model of Distributed Information Retrieval Systems for Internet Applications

随着Internet上的用户急剧增加 ,如何快速地处理查询请求已成为面向Internet的信息检索系统急需解决的问题 .文中提出了一个面向Internet的分布式信息检索系统模型 (Fastresponse) .在该系统模型中 ,针对当前分布式信息检索系统在可获得性、可靠性、可扩展性、查询服务智能性等方面存在的一些问题 ,我们提出了对服务器和文档进行两级调度 ;系统需要时动态招募备份服务器 ;对子服务器群中和子服务器群间的服务器进行两级监控 ;根据用户的要求自动重组服务器等一些新的管理策略 ,有效地解决了上述方面存在的问题 ,从而可以更好地处理用户的查询请求 .

With the rapid increase of Internet users,it is more and more important for Internet information retrieval systems to process large amount of users' query efficiently and rapidly.In this paper,we propose a model,Fastresponse,of distributed information retrieval systems for Internet applications.Currently,the distributed information retrieval systems on the Web have problems in the aspects of availability,reliability,scalability and intelligent query service.In Fastresponse,we propose some methods for the management of Web server to resolve the problems above.Thus,we can better process the users' query.We describe the features of Fastresponse.The system adopts a two-level scheduling policy to arrange the works of servers and documents.The system uses the standby servers that are dynamically selected when the system requires.The system monitors the running servers' states in the same cluster and among the clusters.The system automatically regroups the servers according to the use's request.