The Role of Semantic Locality in Hierarchical Distributed Dynamic Indexing

摘要

The global growth in popularity of the World Wide Web has been enabled in part by the availability of browser-based search tools, which in turn have led to an increased demand for indexing techniques and technologies. This explosive growth is evidenced by the rapid expansion in the number and size of digital collections of documents. Simultaneously, fully automatic content-based techniques of indexing have been under development at a number of institutions. The time is thus ripe for the development of scalable knowledge management systems capable of handling extremely large textual collections distributed across multiple repositories. This paper introduces a technique for identifying regions of semantic locality within a collection of documents represented as a semantic network. Semantic locality imbues semantics to statistically based measures of graph connectedness. Such semantics become the basis for several applications in scalable knowledge management.