Digital Library

cab1

 
Title:      PARAMETER ESTIMATION BASED ON EM ALGORITHM FOR ANTI-FOLKSONOMICAL ITEM RECOMMENDATION SYSTEM IN SOCIAL BOOKMARKING
Author(s):      Akira Sasaki , Takamichi Miyata , Yoshinori Sakai , Yasuhiro Inazumi , Aki Kobayashi
ISBN:      978-972-8924-97-3
Editors:      Hans Weghorn and Pedro Isaías
Year:      2009
Edition:      V I, 2
Keywords:      Social Bookmarking, Anti-folksonomy, Collaborative Filtering
Type:      Full Paper
First Page:      43
Last Page:      50
Language:      English
Cover:      cover          
Full Contents:      click to dowload Download
Paper Abstract:      Collaborative filtering (CF) is a technique widely used in recommendation systems. However, if we have only a sparse data matrix that consists of users’ preferences, CF performance seriously degrades. We developed a CF-based recommendation system using social bookmarking data to solve the problems data sparsity causes[1]. However, we applied our method only to small data set, and we decided the parameters heuristically. In this study, we developed a parameter estimation method by using an EM-algorithm and evaluated our system using a huge dataset obtained from actual social bookmarking services.
   

Social Media Links

Search

Login