Abstract: The Grid technology has already been applied to long-distance education field.Service discovery is an important aspect in long-distance education system.In accordance with the feature of job in Grid，a series of services should be geographically relevant.A novel method is proposed which utilizes Bloom filter to construct a space-efficient service group filter.Simulation results，along with analysis，show that this scheme is context-aware，can avoid deficiency of the existing schemes and is workable and effective.

Key words: grid, service discovery, Bloom filter

摘要： 网格环境下的远程教育出现对网格服务发现提出了挑战，UDDI上基于关键词和简单分类的服务发现机制已经不能很好满足需要。在分析现有相关研究的基础上，提出了基于Bloom过滤器的网格服务发现机制。通过对网格服务立体性分析，指出可对网格服务发布构建Bloom过滤器，给出一个具有高度空间相关性的服务发现方法。理论分析与实验表明，该方法是有效可行的。

关键词: 网格, 服务发现, Bloom过滤器