A Study of Recommender Systems Based on the Latent Class Model Estimated by Combining Both Evaluation and Purchase Histories  

Oleh:

Oi, Takahiro ; Mikawa, Kenta ; Goto, Masayuki

Jenis: Article from Proceeding
Dalam koleksi: The 14th Asia Pacific Industrial Engineering and Management Systems Conference (APIEMS), 3-6 December 2013 Cebu, Philippines, page 1-10.
Topik: recommender system; collaborative filtering; latent class model; parameter estimation; aspect model
Fulltext: 1230.pdf (863.54KB)

Isi artikel

Recently the recommender system which automatically recommends items for customers has become more important as an efficient web marketing tool.In many electric commerce (EC) sites, the data of customer’s purchase and evaluation histories are stored in databases. Byusing them, the systemestimates user’s preference, and automatically recommends items which are seemed to be preferred but have not been purchased yet. In this study, we focus on the recommender system based on collaborative filtering (CF) with the latent class model. CF recommends items by using a latent class model estimated by purchase or evaluation history data. Considering real purchase activity on EC sites, most of consumers who bought items on an EC site don’t post their evaluation on the site. That meanspurchase history data arestored more than that of evaluation history datain the database. However, most conventional studies of CF used only evaluation history data to learn the model. In this case, the purchase history data was not used to construct a model even though its data size is much larger than evaluation history. It is more desirable to learn a model by using not only evaluation history but also purchase history data to improve the CF accuracy. The purpose of this study is to construct an effective CF model to improve the CF accuracy by formulating the estimation byboth evaluation history data and huge purchase history data which has not been used in previous CF studies. Specifically,we focus on the aspect model which is one of latent class models of CF. We propose the way to estimate its parameters by using both evaluation and purchase history data. To verify the effectiveness of this study, a simulation experiment is conducted by using a bench mark data of recommendation. We show that the prediction accuracy of recommendation is improved.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali