Rating Information-Aided Denoising AutoEncoder for Effective Collaborative Filtering 


Vol. 43,  No. 8, pp. 1357-1367, Aug.  2018
10.7840/kics.2018.43.8.1357


PDF Full-Text
  Abstract

Recommendation systems are the ones that recommend a preferred item based on user’s preference, and typically include a collaborative filtering method. However, collaborative filtering has a shortcoming such that recommendation accuracy is degraded if there is no sufficient rating information assigned by users. In this paper, we introduce a new collaborative filtering method that employs denoising autoencoder which is one of machine learning techniques mainly used for sparse data, and show the improved recommendation accuracy. In the proposed method, to effectively predict ratings, a linear combination of the rating average of a target user and the rating average of a target item is considered. In other words, the rating is predicted by a weighted sum of each rating average and each weight is learned through denoising autoencoder. The performance of our model is demonstrated in terms of precision, recall, F-measure, and nDCG in the top-N recommendation system. When the MovieLens dataset is used, it is verified that the proposed method outperforms the conventional denoising autoencoder-based collaborative filtering by up to 210% in terms of nDCG.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

H. Kim, D. Shin, W. Shin, C. Hwang, "Rating Information-Aided Denoising AutoEncoder for Effective Collaborative Filtering," The Journal of Korean Institute of Communications and Information Sciences, vol. 43, no. 8, pp. 1357-1367, 2018. DOI: 10.7840/kics.2018.43.8.1357.

[ACM Style]

Hyun-Jin Kim, Dong-Jin Shin, Won-Yong Shin, and Changha Hwang. 2018. Rating Information-Aided Denoising AutoEncoder for Effective Collaborative Filtering. The Journal of Korean Institute of Communications and Information Sciences, 43, 8, (2018), 1357-1367. DOI: 10.7840/kics.2018.43.8.1357.

[KICS Style]

Hyun-Jin Kim, Dong-Jin Shin, Won-Yong Shin, Changha Hwang, "Rating Information-Aided Denoising AutoEncoder for Effective Collaborative Filtering," The Journal of Korean Institute of Communications and Information Sciences, vol. 43, no. 8, pp. 1357-1367, 8. 2018. (https://doi.org/10.7840/kics.2018.43.8.1357)

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880