|
Title:
|
LEARNING FROM LOCATION HISTORY AND ITS CONTEXT FOR COLLABORATIVE PEOPLE BEHAVIORAL PREDICTION |
|
Author(s):
|
Hiroki Saito, Tsubasa Takayama |
|
ISBN:
|
978-972-8939-93-9 |
|
Editors:
|
António Palma dos Reis and Ajith P. Abraham |
|
Year:
|
2013 |
|
Edition:
|
Single |
|
Keywords:
|
Web Mining, User Behavior Prediction, Location-based Services, Micro-blog Services. |
|
Type:
|
Poster/Demonstration |
|
First Page:
|
103 |
|
Last Page:
|
106 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
As GPS-enabled mobile devices such as iPhone and Android become popular, many location-based services (LBS) are provided and it becomes quite attractive to users. Since location trajectories of users imply human behaviors and preferences related for users interests, several studies for recording and analyzing peoples activity histories have been done. People can share on the Web not only raw GPS coordinates and time-stamps, but also rich contents such as comments and pictures related to their movement for social blogging. This paper proposes a stochastic model for predicting user behavior by analyzing location trajectories and its context. We evaluated our algorithm using real-world geo-tagged Twitter dataset collected for over six months, and we show that our algorithm is accurate compared with previous statistic-based approach. |
|
|
|
|
|
|
