A heuristic approach to network optimized mapping of a distributed resource discovery architecture

摘要

As the Internet grew into a tremendous collection of networks connecting millions of people, resource discovery became an increasingly important problem. Even though, there are popular search engines that provide indexes for reaching to vast amount of information, there are issues that deserve close attention. First is the directory structure. Due to rapid growth of Internet, a centralized directory can easily become a bottleneck and critical point of failure [6]. Techniques such as replication and mirroring can solve these problems to a certain extent, however, with the introduction of extra machinery, the registry process may become too complicated. Distributed directory structure can be more appealing, however, in such a case, search methods may increase the overall processing and network costs beyond an acceptable level. The second issue is the organization of information. The classical approach of hierarchical organization of information imposes too many restrictions due to strict attributes and deeply nested components. The user may not be leaded to what he wants since the nested components lead to predefined and constant paths. Syntactical relationships that relate two or more concepts belonging to different hierarchies should be developed to solve the problem of the user. However, this in turn further complicates the directory structure.