Paper Abstract and Keywords |
Presentation |
2011-12-16 16:50
A Transition Probability Model for Detecting both Information and Users Satoshi Aoki, Jianwei Zhang, Yukiko Kawai, Yoshinori Kubota, Takuya Osaki, Toyokazu Akiyama (Kyoto Sangyo Univ) DE2011-49 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
PageRank algorithm can achieve a ranking result of a large amount of pages by forming a transition probability matrix consisting of the hyperlinks between pages. However, as for the increasing CGM contents, for example, Blog and Twitter, links are structured by not only between contents and contents, but also between contents and users. Consequently, the conventional ranking model that only analyzes links between pages is difficult to rank pages and users all together at once. We propose a real-time ranking model of simultaneously pages and users, and forming a transition probability matrix consisting of both pages and users by analyzing relationship between pages and users. In this paper, first, we describe the constructing method of a graph based on the relationship of page-page, page-user and user-user, and then we explain the forming method the construction of its corresponding matrix and the ranking calculation method. Finally, we evaluate proposed model by our developed system from experiment. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Information retrieval / User search / Transition probability matrix / Social services / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 111, no. 361, DE2011-49, pp. 53-58, Dec. 2011. |
Paper # |
DE2011-49 |
Date of Issue |
2011-12-09 (DE) |
ISSN |
Print edition: ISSN 0913-5685 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 |
DE2011-49 |
|