Online recommender system for radio station hosting based on information fusion and adaptive tag-aware profiling

摘要: A new implicit feedback recommender system for the interactive radio network FMhost.A collaborative approach paired with dynamic tag-aware profiles or users and radios.An adaptive online learning strategy based on user history information fusion.We compare it an SVD-based technique in terms of precision, recall, NDCG.Our experiments show that fusion-based demonstrates best results. We present a developed Russian FMhost. To our knowledge, is first model associated case study recommending stations hosted by real DJs rather than automatically built streamed playlists. address such problems as cold start, gray sheep, boosting rankings, preference repertoire dynamics, absence explicit feedback, underlying combines user-based personalized from tags listened tracks order to match station profiles. This made possible profiling follows history. proposed algorithms singular value decomposition (SVD) normalized discounted cumulative gain (NDCG) measures; In addition, we give theoretical analysis some useful properties linear combination methods graded ordered sets.