CAP7: Searching and browsing in distributed document collections*

Abstract

This paper describes CAP7, a system for searching and bro- wsing in distributed document (metadata) collections. The system ar- chitecture is similar to Harvest, comprising gatherer components and a retrieval engine; but instead of the limited SOIF data format, we use RDF and XML. The gatherer creates RDF metadata descriptions of collected resources. Before delivering the data to the retrieval engine, the RDF is transformed into valid XML. The query language supported by the retrieval engine is an extension of XQL with weighting and data types with vague predicates. The user interface provides for browsing as well as simple searching.