出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：The cold-start items, especially the New-Items which did not receive any ratings, have negativeimpacts on NMF (Nonnegative Matrix Factorization)-based approaches, particularly the onesthat utilize other information besides the rating matrix. We propose an NMF based approach incollaborative filtering based recommendation systems to handle the New-Items issue. Theproposed approach utilizes the item auxiliary information to impute missing ratings beforeNMF is applied. We study two factors with the imputation: (1) the total number of the imputedratings for each New-Item, and (2) the value and the average of the imputed ratings. To studythe influence of these factors, we divide items into three groups and calculate theirrecommendation errors. Experiments on three different datasets are conducted to examine theproposed approach. The results show that our approach can handle the New-Item's negativeimpact and reduce the recommendation errors for the whole dataset.
