Abstract:

Advanced search tools are important components of the information infrastructure. Different mechanisms, such as: powerful search engines, institutional and collaborative portals, library catalogues, specialized discovery services, indexes, databases and electronic journal aggregation services, are used in order to discover and access informational resources. The paper presents a meta-search tool developed by the authors in order to deliver search results structured according to the specific interests of the users. Meta-search means that for a specific query, several search mechanisms could be applied. Using the clustering process, thematically homogenous groups are building up from the initial list provided by the standard search mechanisms. To identify the clusters, the characteristics of the initial results are used, without any predefined categories.