摘要：Most information in science, engineering andbusiness has been recorded in form of text. Thisinformation can be found online in the World-Wide-Web. One of the major tools to support informationaccess are the search engines which usually useinformation retrieval techniques to rank Web pagesbased on a simple query and an index structurelike the inverted lists. The retrieval models are thebasis for the algorithms that score and rank the Webpages. The focus of this presentation is to showsome inverted lists alternatives, based on buckets, foran information retrieval system. The main interestis how query performance is effected by the indexorganization on a cluster of PCs. The server design iseffected on top of the parallel computing model BulkSynchronous Parallel-BSP.
