作者: Torben Bach Pedersen , François Deliège
DOI:
关键词: Space (commercial competition) 、 Computer science 、 Recommender system 、 Fuzzy logic 、 Perspective (graphical) 、 Data management 、 Similarity (psychology) 、 Bitmap 、 Information retrieval 、 Data structure
摘要: The emergence of music recommendation systems calls for the development new data management technologies able to query vast collections. In this paper, we define fuzzy song sets and an algebra manipulate them. We present a warehouse prototype perform efficient nearest neighbor searches in arbitrary similarity space. Using sets, offers practical solution all musical scenarios provided: comparisons, user preferences feedback. investigate three approaches tackle storage issues sets: tables, arrays bitmaps. Finally, confront theoretical estimates concrete implementation results prove that, from perspective, bitmaps are both effective structure solutions.