Paper Abstract and Keywords |
Presentation |
2018-06-26 09:25
A consideration on the possibility of automatic classifying for anomalous posts on Twitter Ryutaro Ushigome (NICT/Chuo Univ), Takeshi Matsuda (Univ of Nagasaki), Michio Sonoda, Takeshi Takahashi, Mio Suzuki (NICT), Jinhui Chao (Chuo Univ) IA2018-9 ICSS2018-9 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
A lot of information has been posted on the SNS, including not only erroneous information and inappropriate writing, but also writing by machine etc. Techniques for detecting and discovering inappropriate postings in SNS have been studied in the past, but it is not easy to automate this and it is a social problem as well. In this paper, unsupervised learning is conducted on data collected from Twitter, one of SNS. Then the result is analyzed to examine whether posts including misinformation or mechanical postings can be classified. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
SNS / Twitter / K-means / unsupervised machine learning / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 118, no. 109, ICSS2018-9, pp. 55-60, June 2018. |
Paper # |
ICSS2018-9 |
Date of Issue |
2018-06-18 (IA, ICSS) |
ISSN |
Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
IA2018-9 ICSS2018-9 |
|