Home // AFIN 2015, The Seventh International Conference on Advances in Future Internet // View article

Estimating Consumer Inclination for Agricultural Products from Web Browsing History

Authors:
Riki Tatsuta
Sunao Nakanishi
Yusuke Kajiwara
Hiromitsu Shimakawa

Keywords: recommendation; inclination; intention; preference; agriculture; e-commerce site; browsing history.

Abstract:
In recent years, more and more e-commerce sites sell agricultural products. It should be noted that users have preference of taste and intention on purchase of agricultural products in recommendation on e-commerce sites. There are a number of studies on the automatic extraction of preference of taste that users have from contexts on the web. However, there has been no study that tries to automatic extraction of user intentions to purchase agricultural products in on e-shopping site. E-shopping sites without consideration of user intentions are likely to recommend agricultural products far from what users want to buy. Confused users never visit such e-shopping site, again. We propose a model which estimates intention level of users from web contents of e-commerce sites and web browsing history of users. The proposed method refers to the product of topic distribution calculated from Latent Dirichlet Allocation and the browsing time of web pages as the topic attention level. Based on the multivariate logistic regression model, the method constructs an intention level model fromthe topic attention level. An experiment suggests the method can calculate the safety intention level, focusing on price of agricultural products on web pages.

Pages: 26 to 31

Copyright: Copyright (c) IARIA, 2015

Publication date: August 23, 2015

Published in: conference

ISSN: 2308-4340

ISBN: 978-1-61208-428-2

Location: Venice, Italy

Dates: from August 23, 2015 to August 28, 2015