System and method for clustering data objects in a collection

作者: Hinrich Schuetze , James E. Pitkow , Jun Li , Ed H. Chi , Peter L. Pirolli

DOI:

关键词: PopulationVector spaceInformation retrievalAnalogySimilarity (network science)Data objectsCluster analysisComputer scienceData mining

摘要: A system and method for browsing, retrieving, recommending information from a collection uses multi-modal features of the documents in collection, as well an analysis users' prior browsing retrieval behavior. The are premised on various disclosed methods quantitatively representing document vectors multi-dimensional vector spaces, determining similarity between documents, clustering according to those similarities. also rely users user population, users, similarities, visually clusters by analogy documents.