Home // UBICOMM 2019, The Thirteenth International Conference on Mobile Ubiquitous Computing, Systems, Services and Technologies // View article

Predicting User Interests Based on Their Latest Web Activities

Authors:
Takeshi Tsuchiya
Hiroo HIROSE
Tetsuyasu YAMADA
Hirokazu YOSHINAGA
Keiichi KOYANAGI

Keywords: user interests, content prediction

Abstract:
This paper discusses and proposes a method to predict interesting information of users based on their recent behaviors on web. User activities on web aim to investigate and acquire some information through web sites under his/her interests. Therefore, we assume that recent interests of users will be predicted by the analyzing the characteristics of acquired web contents. Our proposal method learns these user's interests based on clicked log of web advertisement by the neural networks, and enables to predict the information by the regression to learned user model. It means that a flexible information service to be constructed using predictions adapted to the user's presence. The evaluations indicate that the method is effective and practical comparison to conventional model which use statistical model for analyzing web activities.

Pages: 24 to 29

Copyright: Copyright (c) IARIA, 2019

Publication date: September 22, 2019

Published in: conference

ISSN: 2308-4278

ISBN: 978-1-61208-736-8

Location: Porto, Portugal

Dates: from September 22, 2019 to September 26, 2019