Home // International Journal On Advances in Internet Technology, volume 11, numbers 1 and 2, 2018 // View article

An Investigation of Users' Actions Expressed in Tweets Submitted by Using Music Player Applications

Authors:
Yasuhiko Watanabe
Kenji Yasuda
Ryo Nishimura
Yoshihiro Okada

Keywords: music player application; music content; behavior based service; Twitter; social media

Abstract:
What users are doing at a certain point in time is important for designing various services and applications in social media, such as targeted advertisement, news recommendation, and real-world analysis. As a result, in this study, we investigated tweets which users submitted when they were listening to music by using music player applications. We collected 2,000 tweets including hashtags generated by music player applications and found about 65% of them were tweets where impressions were described, 15% of them were tweets where reasons why users were listening to music were described, and 10% of them were tweets where actions while listening to music were described. We applied machine learning techniques to detect tweets where two kinds of actions while listening to music, moving to somewhere or going to bed, were described. Furthermore, we examined whether we can detect tweets where two kinds of action phases, start and middle, were described. In both cases, we obtained the high accuracy and precision. The experimental result shows that our method is useful for providing behavior based services and applications in social media.

Pages: 21 to 30

Copyright: Copyright (c) to authors, 2018. Used with permission.

Publication date: June 30, 2018

Published in: journal

ISSN: 1942-2652