DocumentCode :
2888071
Title :
A Regularized Recommendation Algorithm with Probabilistic Sentiment-Ratings
Author :
Peleja, Filipa ; Dias, P. ; Magalhaes, Joao
Author_Institution :
Dept. de Inf., Univ. Nova de Lisboa, Caparica, Portugal
fYear :
2012
fDate :
10-10 Dec. 2012
Firstpage :
701
Lastpage :
708
Abstract :
Recommender systems (RS) provide personalized suggestions based on users´ past behavior and/or similarities between users´ and products´ profiles. Although we observed a high interest in the research community over RS algorithms these commonly overlook users´ opinions. In this paper, we research the inclusion of sentiment knowledge in RS to improve the overall quality of recommendations. In contrast to similar approaches, we propose a matrix factorization with a new factor to regularize probabilistic ratings. A sentiment analysis algorithm implementing a multiple Bernoulli classification computes these probabilistic ratings. The combination of a regularization factor with probabilistic ratings offers a general framework capable of embedding multiple sources into a theoretical well-founded matrix factorization algorithm. Experiments show that with an evaluation on a dataset with 1.7 million reviews we have successfully introduced a novel approach to incorporate on a RS with inferred rating based in a sentiment analysis framework. Also, replacing explicit ratings by probabilistic inferred ratings the RS performance improves, thus, our proposed framework is able to better accommodate the uncertainty of users explicit rating.
Keywords :
matrix algebra; probability; recommender systems; Bernoulli classification; RS; matrix factorization; probabilistic ratings; probabilistic sentiment ratings; recommender systems; regularized recommendation algorithm; research community; sentiment knowledge; Algorithm design and analysis; Classification algorithms; Equations; Matrix decomposition; Motion pictures; Probabilistic logic; Vectors; opinion mining; recommendation systems; sentiment analysis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Data Mining Workshops (ICDMW), 2012 IEEE 12th International Conference on
Conference_Location :
Brussels
Print_ISBN :
978-1-4673-5164-5
Type :
conf
DOI :
10.1109/ICDMW.2012.113
Filename :
6406508
Link To Document :
بازگشت