2013/05/15 by Naseem Biadsy, Lior Rokach, Biadsy, Naseem +3
Computer Science · #Advanced Graph Neural Networks #FOS: Computer and information sciences #Information Retrieval (cs.IR) #Information Retrieval and Search Behavior #Machine Learning (cs.LG) #Recommender Systems and Techniques #Text and Document Classification Technologies
paper · pdf · doi:10.48550/arxiv.1305.3384
openalex publication_date 2013/05/15 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In this paper we present a new approach to content-based transfer learning\nfor solving the data sparsity problem in cases when the users' preferences in\nthe target domain are either scarce or unavailable, but the necessary\ninformation on the preferences exists in another domain. We show that training\na system to use such information across domains can produce better performance.\nSpecifically, we represent users' behavior patterns based on topological graph\nstructures. Each behavior pattern represents the behavior of a set of users,\nwhen the users' behavior is defined as the items they rated and the items'\nrating values. In the next step we find a correlation between behavior patterns\nin the source domain and behavior patterns in the target domain. This mapping\nis considered a bridge between the two domains. Based on the correlation and\ncontent-attributes of the items, we train a machine learning model to predict\nusers' ratings in the target domain. When we compare our approach to the\npopularity approach and KNN-cross-domain on a real world dataset, the results\nshow that on an average of 83% of the cases our approach outperforms both\nmethods.\n