Mittag, KristinHinneburg, AlexanderFähnrich, Klaus-PeterFranczyk, Bogdan2019-01-112019-01-112010978-3-88579-270-3https://dl.gi.de/handle/20.500.12116/19318Information needs like searching scientific literature that involve high recall rates are difficult to satisfy with ad hoc keyword search. We propose to state queries implicitly by the specification of a set of query documents. The result of such a query is a set of answer documents that are ranked within the answer set. We describe efficient techniques to process such queries. Preliminary experiments using data from the TREC Genomics track 2005 are reported.enRanked Set Search in Medline DocumentsText/Conference Paper1617-5468