Login
Register
Reset Password
Publish & Distribute
Publishing Solutions
Distribution Solutions
Subjects
Architecture and Design
Arts
Business and Economics
Chemistry
Classical and Ancient Near Eastern Studies
Computer Sciences
Cultural Studies
Engineering
General Interest
Geosciences
History
Industrial Chemistry
Jewish Studies
Law
Library and Information Science, Book Studies
Life Sciences
Linguistics and Semiotics
Literary Studies
Materials Sciences
Mathematics
Medicine
Music
Pharmacy
Philosophy
Physics
Social Sciences
Sports and Recreation
Theology and Religion
Publications
Journals
Books
Proceedings
Publishers
Blog
Contact
Search
EUR
USD
GBP
English
English
Deutsch
Polski
Español
Français
Italiano
Cart
Home
Journals
Journal of Data and Information Science
Volume 1 (2016): Issue 1 (February 2016)
Open Access
Understanding the Correlations between Social Attention and Topic Trends of Scientific Publications
Xianlei Dong
Xianlei Dong
,
Jian Xu
Jian Xu
,
Ying Ding
Ying Ding
,
Chenwei Zhang
Chenwei Zhang
,
Kunpeng Zhang
Kunpeng Zhang
and
Min Song
Min Song
| Sep 01, 2017
Journal of Data and Information Science
Volume 1 (2016): Issue 1 (February 2016)
About this article
Previous Article
Next Article
Abstract
Article
Figures & Tables
References
Authors
Articles in this Issue
Preview
PDF
Cite
Share
Article Category:
Research Paper
Published Online:
Sep 01, 2017
Page range:
28 - 49
Received:
Jan 18, 2016
Accepted:
Feb 27, 2016
DOI:
https://doi.org/10.20309/jdis.201604
Keywords
Social media
,
Publication topic trends
,
Correlation
,
State-space model
,
Variable selection
,
Nowcasting
© 2016 Xianlei Dong, Jian Xu, Ying Ding, Chenwei Zhang, Kunpeng Zhang, Min Song
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
Figure 1
Google Trends graph showing (a) weekly search popularity of “obesity,” (b) monthly search popularity of “14,” and (c) average search trend of all the queries related to topic “child obesity.”
Figure 2
The overall framework of the methodology, where (1) p(z|d) denotes the probability that document d belongs to topic z; (2) β denotes the keywords’ effects on topics, that is, the coefficients of X; (3) spike γ can make most of the coefficients of X zeros, which ensures that the stepwise regression process will run correctly; and (4) Y – Z*α (regression component) refers to publication data with the time-series component, where tendency and seasonal components are not included.
Figure 3
The monthly number of publications on (a) “child obesity” and (b) “diabetes” over time.
Figure 4
Trends of topics (a) “child obesity” and (b) “diabetes.” The x-axis represents time from January 2004 to January 2013; the y-axis represents the number of publications within a month on “child obesity” and “diabetes,” respectively. Growth rate of topics (c) “child obesity” and (d) “diabetes.” The x-axis represents time from January 2004 to January 2013; the y-axis represents the growth rate of publications within a month on “child obesity” and “diabetes,” respectively.
Figure 5
Seasonal effect for topics (a) “child obesity” and (b) “diabetes” from January 2008 to January 2013.
Figure 6
Regression components for obesity topics (a) “child obesity” and (b) “diabetes” over time.