Research on the characteristics of high-cited papers on pandemics

from the perspective of alternative metrology

Zongfeng Zou
Shanghai University
Xi-yun Sun (  [email protected] )
Shanghai University
Lu Nan
Shanghai University
Wei Shao
Shanghai University
Jingwen Wu
Shanghai University

Research Article

Keywords: Alternative Metrics, Pandemics, Academic Papers, Social Media, Twitter

Posted Date: November 30th, 2023

DOI: https://doi.org/10.21203/rs.3.rs-3610227/v1

License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License

Additional Declarations: No competing interests reported.

Page 1/19
Abstract
[Background] At present, an increasing number of researchers have studied pandemics and published papers to help people better cope
with pandemics.

[Purpose/significance] This study explores the value of Altmetrics to facilitate academic responses to pandemics on the web and helps
public health organizations design communication strategies through an analysis of the dissemination characteristics of high-cited
pandemic papers on social media. To excavate the potential of pandemic papers from the perspective of alternative measurement is also
beneficial to more objectively and effectively evaluate the social influence of those papers.

[Method/process] Correlation analysis was carried out between traditional bibliometric indicators and various Altmetrics indicators. We can
also more clearly understand the characteristics of academic dissemination on Twitter and Mendeley platforms by using visual analysis
and descriptive statistical analysis methods.

[Result/Conclusion] Our findings conclude that online media has become an important tool for disseminating scientific knowledge related
to global pandemics. The Altmetric indicators are significantly different from the citation indicator, which can express the impact of the
high-cited pandemic papers from a new perspective. In conclusion, Altmetrics will be a perfect supplement to the quantitative evaluation of
such papers.

1. Introduction
Human history has seen the occurrence of global pandemics such as cholera, influenza A (H1N1), Ebola virus, SARS, and COVID-19, which
seriously threatened the survival of mankind. The new round of the epidemic has become a devastating disaster, and it is also changing
people's interactive behavior, including scientific communication and academic dissemination. Due to COVID-19 is a highly infectious
disease ,resulting in the implementation of mass quarantines around the world.Social media platforms (e.g. ,Twitter, blogs, Facebook,
Weibo, Wikipedia) are all academic communication during the pandemic, having become one of the most important information channels
for comprehensive, real-time, non-physical communication between the world and users (Chong et al,2021). At the same time, as the carrier
of research findings and scientific exchanges between experts and scholars during the pandemic, the paper is also the voice of researchers
in helping the society respond to and solve the problems people are facing. There are as many as 280,000 papers about COVID-19 Only on
the Web of Science platform, and the growing number of pandemic papers makes it necessary to measure the expressiveness of these on
social media(Li et al,2021;Rehemtula et al,2014). There are many benefits of using social media platforms during a crisis, such as
disseminating information and providing support to the public (Edakar et al,2021). In this way, the general public in non-academic circles
can also better understand the information of the pandemic, grasp the characteristics of the pandemic, and eliminate some fears about the
global epidemic. Experts and scholars are willing to share and discuss their academic research results with the public on social platforms,
and even guide the general public to read and even disseminate important papers during the pandemic, so as to expand the value of their
research to a greater extent, which also reflects the era of big data requirements for researchers.

During the epidemic of SARS in 2002 and influenza A (H1N1) in 2009, in order to speed up the time of knowledge dissemination and make
the scientific research findings available to the society more quickly, a preprint system has been implemented, that is, scientific publication
before peer-reviewed articles (Roelvis et al,2020). With the rapid development of information technology and the advent of the era of big
data, social media, news media, and online reference management software have built an open and free academic exchange ecosystem
(Zhao,2020), which further accelerates the time for knowledge dissemination. At the same time, due to the advent of the Web2.0 era, the
alienation of citations caused by the Matthew effect, the development of the open access of academic papers, and online academic
exchanges are more and more respected by experts and scholars. a doctoral student from the Department of Library and Information at
the University of North Carolina, Priem Jason ,first proposed the concept of Altmetrics on his twitter in 2010, and then Priem J, Taraborelli
D, etc. published the article "Altmetrics: A Manifesto" and formally proposed the word Altmetrics, defining it as an activity based on network
tools and environments on the measurement of academic influence(Priem J et al,2010), which refers to the measurement of all other
information sources except citation data on social networks, including various information behaviors of people"collection, comment, share,
bookmark, and forward". All can be used as measurement indicators of Altmetrics (Yao,2017;Lane J,2010). Compared with the traditional
impact evaluation method of scientific research papers, Altmetrics has a more efficient, timely, diverse and comprehensive impact
coverage, not only considering the academic impact of the paper, but also the social impact and economic impact of the
paper(Wu,2015;Finbar et al,2013;Dinamore,2014;Yang et al,2015).

This research investigates global pandemic papers shared on social media outlets while focusing on measuring the social media attention
given to global pandemics papers indexed in the WoS(Web of Science) database. The Altmetrics.com website was selected to collect

Page 2/19
statistics on the Altmetrics indicators data of highly cited papers on the global pandemic of WoS and the characteristic data of some
Altmetrics indicators, Then performing coverage test on these Altmetrics indicators data and do correlation analysis with the citation
frequency and usage frequency of these pandemic papers. How does the AAS (Altmetric Attention Score) perform on these pandemic
papers? Which Altmetrics indicators can be used as a supplement to traditional indicators to measure the academic influence of a paper ?
It also lays the foundation for a more scientific, comprehensive and effective evaluation of the impact of high-cited papers on the
pandemic. Previous research has proved that Twitter is the most frequently used social media software for academic communication and
the highest mass participation, so what are the characteristics on the geographical distribution and the Occupation types of users who
participate in discussions on Twitter for high- cited papers on pandemics. Mendeley is the literature management software with the highest
coverage rate and the strongest correlation with the citation frequency and impact factor of the paper. What are the characteristics of the
education level of its readers? Does the higher the education level, the more readers participate in the discussion? What are the
characteristics of the subject classification of Mendeley readers? Through the research on the altmetrics indicators of the high-cited papers
of pandemics, aiming to investigate the characteristics and laws of academic communication research in social media under emergency
pandemics. At the same time, from the perspective of alternative measurement, we will explore the potential of high-cited papers in the
global pandemic, and lay the groundwork for predicting future citations.

2. Literature
Since the concept of Altmetrics was formally proposed in 2010, it has caused heated discussions in the global academic circles. In recent
years, researchers mainly focus on the sources and coverage rate of Altmetrics indicators, the correlation between Altmetrics indicators
and citations, the application of Altmetrics and the use of Altmetrics to predict citations as an early indicator of scientific impact. Since the
outbreak of COVID-19 at the end of 2019, a small number of scholars have begun to use Altmetrics, a new bibliometric method, to measure
the impact and spread characteristics of new coronavirus-related papers in social networks.

2.1 The sources and coverage rate of Altmetrics indicators

The sources and coverage rate of Altmetrics indicators has always been the fundamental issue of Altmetrics research. The current
diversity of Altmetrics platforms and indicators means that it is difficult to find a balance between them (Wang,2021). Hammarfelt selects
the literature in the field of social sciences for research. The empirical results show that Mendeley, Twitter, and Facebook have higher
coverage, while the coverage of Video, Policy, and Wikipedia is lower (Hammarfelt,2014). SK Banshal found that medicine and biology
account for more than 50% of all instances of Altmetrics. The coverage rate of different Altmetric platforms also varies, and the overall
coverage of Twitter and Mendeley is much higher than that of Facebook and News (Banshal et al,2019). Alperin's research found that the
coverage rate of indicators based on social networking site indicators and academic management platforms such as Twitter, Facebook,
Mendeley, and Blogs is relatively high, and the coverage rate of other types of indicators is relatively low (Alperin,2015). By analyzing API
data, Robinson found that Twitter and Mendeley have the highest coverage rate among many indicators of Altmetrics, and the coverage
rate of other indicators is quite different in comparison (Robinson,2014). Some scholars study the differences in the coverage of Altmetrics
indicators in different disciplines and found that the coverage of Altmetrics indicators in different disciplines is quite different. The
coverage rate of Citeulike indicators in computer science is higher, and the coverage rate of Twitter indicators in social sciences is higher
(Balke,2015) . Huang Xiao analyzed the Altmetrics indicators coverage of 28,483 high-cited papers in different disciplines, and found the
coverage of computer science and mathematics Mendeley readers is relatively low (Huang et al,2019). Fang Zhichao took more than 2
million papers in different disciplines , and found that papers in the fields of biomedicine and health sciences, humanities and social
sciences have the highest Twitter data coverage (Fang et al,2019). Some scholars have also shown that the coverage of altmetric
indicators in some journals is generally low. Only about 15% of the 500,229 papers selected by Wouters P had an Altmetrics score, and
papers from the social sciences, humanities, medicine and life sciences had high Altmetrics coverage, indicating that papers in these areas
are more popular in social media . (Costas et al,2015). The same is true for Thelwall., who found that PubMed articles had low coverage
(significantly below 20%) for all altmetric sources, except Twitter. Of course, this is due to the fact that their research occurred earlier (the
concept of Altmetric was proposed for a short period of time), whether scholars, researchers or the general public have not yet developed
the habit of sharing and discussing scientific research achievements on online media (Thelwall et al,2013).

2.2 Correlation between Altmetrics and citations

After many scholars have done a lot of research on the concept, definition, sources, coverage of Altmetrics indicators, Altmetrics tools and
other basic issues. some scholars have begun to conduct further research from the perspective of the relationship between Altmetrics

Page 3/19
indicators and citations. Paul Wouters made an extensive comparison of Altmetrics and citations from a multidisciplinary perspective,
confirming the positive correlation between Altmetrics and citations drawn from many previous studies. But the correlation is weak,
indicating that the Altmetrics indicators reflect a different impact compared to the citation (Costas et al,2015). Thelwall concluded the
relationship between Facebook mentions and citations varied by field of study. General medicine and internal medicine, cardiovascular
systems and hematology, and biological sciences have the strongest correlations between Facebook mentions and citations, likely due to
the high citation frequency and popularity of papers in these fields (Mohammadi et al,2020) . Haustein confirmed that the correlation
between Twitter mentions and citations is low, implying that metrics based on Twitter mentions differ from citations-based metrics on
papers. And proposed a framework based on Altmetrics metric coverage and the correlation of Twitter mention metrics and citation
frequency metrics to facilitate the evaluation of paper influence based on new metrics on social media(Haustein et al,2013). Seong
obtained through experiments that the correlation coefficient between citation frequency and AAS is 0.246, indicating that the correlation
between the two is weak(Seong et al,2017). Chen Huafang proved that social media and news media indicators (such as Twitter and
News) are not related to most traditional measurement indicators, and can supplement the traditional citation evaluation system to expand
the measurement scope of the influence of academic achievements.literature management platforms (such as Mendeley) and traditional
measurement indicators are related, which can reflect the value of the paper to a certain extent and predict the citation frequency (Chen et
al,2021). Zhao Wanxin's research on the correlation between citations of Chinese high- cited papers and Altmetrics indicators.Although
social media lags of about a year, it still plays a positive role in promoting the dissemination and dissemination of papers(Zhao et
al,2020). The 7946 papers collected by You Qingbin, analyzing the correlation between the Altmetrics indicators of these papers and the
citations. the correlation coefficient between Mendeley readers and citations is up to 0.450, the correlation coefficient between CiteULike,
Twitter, Facebook, F10004 indicators and citations is between 0.2-0.3, and the correlation coefficient between Blogs, Google+ and citations
is close to 0.2, while the correlation coefficient between News, Video, Pinners, Redditors and citations is less than 0.1. It shows that the four
indicators of News, Video, Pinners, and Redditors have little effect on the academic influence of papers (You et al,2014).

2.3 COVID-19 and Altmetrics

In recent years, academic exchanges on online media have been increasingly popular among scholars, researchers and even the general
public, which also prompted many researchers to study the characteristics of the dissemination of the pandemic papers and the role these
social media platforms have played during the pandemic. Especially for COVID-19, the papers of COVID-19 have excellent altermative
measurement coverage, and scholars have tried to explore the impact of social media platforms on COVID-19 information dissemination
and research (Edakar et al,2021). Vysakh attempted to measure the level of attention given to COVID-19 research on popular social media
platforms. The findings suggest that while social media platforms do not give precise measures, such as the impact provided by citations,
social media metrics can be used to describe the social impact of academic papers and point to the validity of research on COVID-19
(Vysakh et al,2020). Chong studied Twitter as an information carrier and information sharing during pandemics, Defining the Twitter
network as an information channel containing an information source with embedded messages. The results show that Information
carriers, namely information channels, information sources and information dissemination are interrelated, forming an information
organism, which Can help public health organizations design communication strategies that often enable rapid decision-making in
epidemic situations to manage emergency needs(Chong et al,2021). Ji Yoon Moon studied the 100 papers most widely circulated in online
media. People use social media not only to express personal emotions and opinions, but also to find information about the pandemic,
illustrating the key role of social media in disseminating reliable information to the public from national health authorities, public health
experts and scientists. (Moon J et al,2021). Fabiano analysed the COVID-19 papers by comparing the spread on Twitter and citation rates,
social media remains a unique and powerful tool for widespread dissemination of academic publications, further highlighting the
importance of Altmetrics in times of rapid research expansion, such as the COVID-19 pandemic, to localize highly disseminated papers
(Fabiano et al,2020). Tornberg used Altmetric and PlumX metrics to assess the spread of COVID-19 articles on social media, and research
shows that the use of social media helps spread COVID-19 information to the general public and health professionals. In the case of
current COVID-19 articles, Altmetric and PlumX metrics should be used to complement traditional citation counts when assessing the
spread and impact of a COVID-19 paper (Tornberg et al,2020). Roy found that social media platforms are seen as a revolutionary means to
eliminate inter-racial inequalities in access to health promotion and information dissemination in india (Roy et al,2021).

3. Data and metrics

3.1 Data acquisition
This paper integrates two data providers, Altmetric.com and Web of Science (WOS). As one of the important platforms for social media
data at present, Altmetric.com has reliable data sources. It introduces new social media indicators (such as traditional news and policy
Page 4/19
media mentions) that other platforms do not have, and the indicators cover a wide range (Wu et al,2016). Enter the Altmetric.com website
and input the DOI number of each paper from the Altmetric Explorer to get the paper Altmetrics data. WOS is an authoritative, powerful and
comprehensive scientific and technical literature indexing tool. This study will search the global pandemic literature in WoS with the
keywords "global pandemic" and "global infectious diseases" without any restrictions on journals, languages, publication dates, output
titles, output types or academic identifiers . Searches in WOS and Altmetric Explorer were conducted from November 25th to November
28th, 2020, to avoid changes in paper AAS as much as possible. In order to determine the most matching publications, select high-cited
papers, and then input the search results into Excel in descending order of citation frequency, determineing 1,327 papers on the topic of
"global pandemic" and 733 papers on the topic of "global infectious diseases".

3.2 Data preprocessing

When searching the Web of science site with "global pandemic" and "global infectious diseases" as the subject keywords, there will be
some duplicates in these two topics, and 227 duplicates will be deleted. 45 of the remaining 1833 papers do not have a DOI number, that
is, the Altmetric data of these 45 papers cannot be queried, and 1778 papers remain after deletion. In addition, there are 48 papers with the
same content but slightly different titles after inspection, so 1730 papers remain after deletion. According to Newman's rule of thumb for
determining the sample size (Khosla I,2021), it can be seen that statistics for small samples, the sampling ratio needs to be about 30% to
achieve a relatively high accuracy, according to which the reliability and validity of this study can be guaranteed (Shi et al,2021). Among
the remaining 1730 papers, 1398 papers have AAS, and 332 papers do not have AAS, that is, these 332 papers have high citation
frequency but no online social platform data. In order to ensure the rigor and accuracy of the study, the data acquisition and screening
process was repeated twice by one investigator, and the same results were obtained. Recording the citation frequency, usage, Altmetrics
attention scores (AAS), main Altmetrics indicators data, and Altmetrics indicators characteristic data of 1398 papers. Figure 1 has been an
paper with the highest AAS on the Altmetric.com website so far, and it clearly shows the specific total score and the value of each
Altmetrics indicators.

3.3 Indicator selection

After completing data acquisition and data preprocessing, this study will select indicators respectively. The traditional paper measurement
indicators include the citation frequency of papers in WOS, the number of times of use in WOS from 2013 to the present, and the number
of times of use in WOS in the last 180 days. Altmetrics indicators include AAS, News mentions, Blogs Mentions, Policy Mentions, Twitter
Mentions, Patent Citations, Weibo Mentions, Facebook Mentions, Wikipedia Citations, Google+ Mentions, Reddit Mentions, Vedio Mentions,
Mendeley readers, Dimensions readers and CiteULike mentions, a total of 15 indicators.

4. Results and Analysis

4.1 Analysis of Research Topics in Pandemic High-Cited Papers
In this study, the research topics of the high-cited papers on pandemics were first discussed. In order to visualize the research topics of
these, the dataset was imported into the VOSviewer software. Cluster analysis was performed on the keywords (more than 20 times)
studied in the titles and abstracts of 1398 papers (Fig. 2).Size of the clusters in the figure indicates the number of occurrences, the color of
the clusters indicates the similarity of the research contents, and the distance indicates the closeness of the research relationship (Ou et
al,2020). The cluster map clearly shows the three main themes of this research dataset: coronavirus characteristics, symptoms and
outcomes, epidemiological research and methods, which are related to the blowout paper explosion of scholars and researchers since the
outbreak of the 2019 coronavirus. Although previous studies have concluded that there is a 1–3 year delay in the citation time of general
papers, the citation accumulation speed of papers related to the pandemic virus does not have such a delay. These three main research
directions that are closely related to human beings or affect the importance of human life and death. However, it can also be found from
the correlation of the research topics in the figure that the current high-cited papers on pandemics are not strongly related to social media,
which means that at this stage there are few scholars and researchers who Analysis and research on the characteristics of pandemics on
the network media. Therefore, the characteristics of pandemics high-cited papers on online media are also worthy of further exploration,
and can have a positive impact on society response to the COVID-19 and other global pandemics in the future.

4.2 AAS Score Ranking Distribution of Pandemic High-Cited Papers

Page 5/19
Figure 3 shows the distribution of the AAS scores of high-cited papers in pandemic (excluding 17 papers without ranking data, the
remaining 1381 papers). To date, Altmetric.com has tracked 19,864,708 research papers. Series 1 refers to the AAS score ranking of
research data in all papers, and Series 2 refers to the AAS score ranking of papers in the same period. The data shows the remaining 1381
essays, 74.4% of the papers are in the top 5% of all papers tracked by Altmetric.com and 88.9% are in the top 10% of all papers tracked by
Altmetric.com. There are 50 papers in the top 100 papers in the AAS ranking of the same period, and 239 papers in the top 1000 papers in
the same period. This data confirms that high-cited papers on pandemics have received great attention in social media, news media, and
online reference management software, so the next step is to further verify the coverage of alternative metrics through the Altmetrics
indicators coverage test.

4.3 Altmetrics indicators coverage test

In the process of data preprocessing, there are a total of 1730 papers left without duplication and no DOI, which are recorded as the total
data set, and recorded as N. There are 1398 literatures with non-zero AAS values ​in the total data set, which is called the Altmetrics data
set, denoted as A, and A is the main research object in this study. Table 1 shows the coverage rates of various Altmetrics indicators in the
research dataset A. The indicators with higher coverage rates can reflect their value and influence in evaluation to a certain extent (Li et
al,2019). It can be obtained that the mendeley indicators, the Dimensions indicators, and the Twitters indicators have the highest coverage
rates, reaching 80.81%, 80.58%, and 79.60%, all exceeding 70%. This is in line with the conclusions of most previous studies. Twitter is the
most popular social media software and Mendeley is the most popular online reference management tool, and they are the most effective
source of metrics for further research by Altmetrics. In addition, the coverage rates of News, Blogs, and Facebook indicators reached
59.88%, 48.38%, and 43.87% respectively, and the coverage rates of these three indicators were between 40% and 60%, indicating that the
influence of high-cited papers in the field of global pandemics is concentrated in six indicators: Mendeley, Dimensions, Twitter, News,
Blogs, and Facebook. Followed by the Policy, Wikipedia, Redditors, and Vedio indicators coverage rates were 31.04%, 21.33%, 15.03%, and
10.81% respectively, and the coverage rates of these four indicators were between 10% and 35%. The coverage rates of Google+, CiteULike,
Patent, and Weibo indicators are 6.71%, 6.13%, 4.74%, and 0.40% respectively, all lower than 7%. Since the coverage rate of these four
Altmetrics indicators is too low, which means that the academic papers in the network media represented by these indicators are less
mentioned, these indicators cannot provide a good reference for further research. (Huang et al,2019). Therefore, the following research only
selects ten Altmetrics indicators with a coverage rate greater than 10%.

Table 1
The meaning and coverage rate of Altmetrics indicators
Number Indicators papers(A = 1398) coverage rate(N = 1730) Indicators menning

1 Twitters 1377 79.60% Twitter mentions

2 Facebook 759 43.87% Facebook mentions

3 Mendeley 1398 80.81% Mendeley readers

4 Dimensions 1394 80.58% Dimension readers

5 News 1036 59.88% New mentions

6 Blogs 837 48.38% Blog mentions

7 Policy 537 31.04% Policy mentions

8 Weibo 7 0.40% Weibo mentions

9 Wikipedia 369 21.33% Wikipedia mentions

10 Redditors 260 15.03% Reddit mentions

11 Google+ 116 6.71% Google Plus mentions

12 CiteULike 106 6.13% CiteULike mentions

13 Vedio 182 10.81% Video mentions

14 Patent 82 4.74% Patent mentions

4.4 Indicators correlation test

Page 6/19
The correlation test judges the relationship between the indicators through the correlation coefficient and the significance level. It is a
statistic to study whether there is some kind of dependency between the observed values ​of each variable. A statistical method to explore
the direction and degree of correlation of variables with specific dependencies. It is expressed by the correlation coefficient. The larger the
absolute terms of the correlation coefficient, the stronger the correlation, indicating that as one variable increases or decreases, another
variable increases or decreases significantly. On the contrary, the weaker the correlation, which means that with the increase or decrease of
one variable, there is no significant increase or decrease of the other variables, indicating that the variable is a factor that affects the other
variables, but it is not the only factor (Tan et al,2007). Which Altmetrics indicators can well represent the academic impact of pandemic
papers and can be used as a supplement to traditional indicators such as citation frequency and impact factor? and which Altmetrics
indicators can only be used to measure the impact of other aspects of papers? Both are verified by correlation analysis between traditional
indicators and Altmetrics indicators. This process also lays the foundation for a more scientific, comprehensive and effective evaluation of
the impact of high-cited papers on pandemics. Altmetrics indicators applied to papers, scholars, and institutional influence evaluation
research are mostly carried out based on this method. Only the Altmetrics indicators that are significantly correlated with traditional
citation indicators are more suitable for inclusion in the academic impact evaluation system of pandemic papers. In this study, the

Table 2
Single-sample Kolmogorov-Smirnov detection of cited metrics
Comparison items Citaions AAS Twitters Facebook Mendeley news blogs Vedio policy Wikipedia

Normal Average 226.02 387.60 304.19 4.96 611.63 22.99 2.91 0.25 0.85 1.12
distribution
Standard 438.54 973.73 1128.25 38.30 865.78 52.59 5.95 0.87 1.731 9.751
parameters
deviations

K-S test 0.308 0.347 0.394 0.448 0.251 0.331 0.312 0.479 0.312 0.454

Progressive significance 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

one-sample Kolmogorov-Smirnov test was used to verify the normality of various indicators. All the indicators in this study were tested
successively, and the asymptotic significance results of all indicators were 0.000, less than 0.05, Therefore, these indicators do not
conform to the normal distribution. Due to the large number of indicators, the space is limited, so the test results are listed in part (Table 2).

Based on this, this study uses the Spearman correlation coefficient method to analyze the correlation between the traditional indicators of
pandemic papers and the Altmetrics indicators. Table 3 shows the results of the correlation analysis between the traditional indicators and
AAS. It can be seen that the correlation coefficient between the citation frequency and the number of use since 2013 is the highest,
reaching 0.619, indicating that there is a strong correlation between these two indicators, that is, the number of high-cited pandemic papers
in WOS from 2013 to the present can also be a good indicator of the academic influence of the papers. The academic influence of the
paper. The correlation coefficient between the citation frequency and the number of uses in the last 180 days is only 0.171, indicating that
the number of uses in the last 180 days can not make a good prediction of the academic influence of the paper. The correlation coefficient
between AAS and citation frequency is 0.415, which is a moderate correlation, indicating that the AAS indicator can reflect the academic
influence of papers to a certain extent, and also measure the influence of high-cited pandemic papers in other aspects. The AAS of the first-
cited paper in the research data is only 943, ranking the 142th in the AAS ranking of papers, and only one paper in the top ten cited papers
ranks in the top ten in the AAS ranking, two papers Ranked in the top 20 of the AAS rankings. Similarly, the paper with the first AAS in the
research data is cited only 87 times and the 788th of all papers. It shows that a high- cited paper with a high academic impact on the
global pandemics does not necessarily have a large social influence. In social networks, the most discussed and mentioned high-cited
papers on the global pandemic are not necessarily the greatest academic impact. But papers with high AAS have excellent social influence,
which conveys the current status of some scientific research or investigations to the public, and is also the scientific knowledge that the
public urgently needs and pays attention to. That is to say, when users of various platforms in the network conduct academic activities on
the platform, they do not only look at whether the paper has many citations and the academic influence is large. Other factors may be
more dominant.

Page 7/19
 Table 3
Correlation coefficient between traditional indicators and AAS
Indicators Citation Last 180 days 2013 to present AAS

usage count usage count

Citation 1.000 .171** .619** .415**

Last 180 days .171** 1.000 .630** .138**

usage count

2013 to present .619** .630** 1.000 .158**

usage count

AAS .415** .138** .158** 1.000

Note: **Statistically significant at p = 0.01

Next, Spearman correlation analysis was performed on citations and other indicators of Altmetrics. The results show that all Altmetrics
metrics are positively correlated with citations. The highest correlation between Altmetrics indicators and citations is Dimensions, and the
correlation coefficient is 0.972, indicating that the Dimensions indicator has a strong explanatory power for the citation frequency of high-
cited pandemic papers, and can almost reflect the academic influence of the paper as well as the citation frequency. The Mendeley
indicator also has a high correlation with the citation frequency, with a correlation coefficient of 0.775, which is consistent with the
previous conclusions drawn by Chen Huafang(Chen et al,2021) and You Qingbin (You et al,2014). The Mendeley indicator of the literature
management platform has a high correlation with the traditional measurement indicator, which can well reflect the value of the paper. The
three indicators of Facebook, blogs, and Policy are moderately related to the citation frequency, and the correlation coefficients are all
around 0.45, which is equal to the AAS above, indicating that these three indicators can reflect the academic influence of the paper to a
certain extent, and also measure the Other aspects of the impact of high-cited pandemic papers. The correlation coefficient between
Twitters, News, and Wikipedia indicators and citation frequency is 0.3–0.4, and the correlation is low. In particular, although the Twitters
indicator has a high coverage rate, it has a good foundation for mass dissemination of scientific papers and online academic exchanges.
Metrics based on Twitter mentions differ from citation-based metrics on the impact of papers, which is consistent with the conclusion of
Haustein S. (Haustein et al,2013).The Redditors and Vedio indicators have the lowest correlation with citation frequency, and the
correlation coefficient is only about 0.2, which can almost only be used to measure the social influence of papers.

Table 4
Correlation coefficient between citation frequency and various indicators of Altmetrics
Indicators Citation Twee Face Mend Dimen news blogs policy Wikip Redd Vedio

ters book eley sions edia itors

Citation 1.000 .296** .441** .775** .972** .379** .443** .445** .398** .189** .209**

Twitters .296** 1.000 .608** .306** .326** .580** .546** .260** .288** .477** .319**

Facebook .441** .608** 1.000 .378** .442** .451** .461** .315** .327** .366** .283**

Mendeley .775** .306** .378** 1.000 .787** .366** .421** .398** .346** .197** .260**

Dimensions .972** .326** .442** .787** 1.000 .407** .470** .443** .395** .221** .227**

news .379** .580** .451** .366** .407** 1.000 .682** .352** .351** .435** .318**

blogs .443** .546** .461** .421** .470** .682** 1.000 .362** .416** .472** .320**

policy .445** .260** .315** .398** .443** .352** .362** 1.000 .298** .164** .161**

Wikipedia .398** .288** .327** .346** .395** .351** .416** .298** 1.000 .262** .247**

Redditors .189** .477** .366** .197** .221** .435** .472** .164** .262** 1.000 .273**

Vedio .209** .319** .283** .260** .227** .318** .320** .161** .247** .273** 1.000

Note: **Statistically significant at p = 0.01

4.5 Feature Analysis of Twitters indicator

Page 8/19
4.5.1 Occupational types of Twitter mentions users
Due to the high coverage rate of Twitter indicator, users on Twitter, whether scholars, researchers or the general public, have gradually
developed the habit of sharing and discussing the scientific papers on the platform. This study begins with a discussion of the
occupational types of users who mention high-cited pandemic papers on Twitter (Table 5), and the percentage of Members of the public,
scientists, Practitioners(doctors, other healthcare professionals), and science communicators (journalists, bloggers, editors). The
percentages of total mentions were 80.1%, 11.7%, 6.1%, 0.2% respectively. It can be seen that general public prefer sharing and discussing
high-cited pandemic papers on Twitter, and the proportion is much higher than users of other Occupational types, This is contrary to our
impression that researchers such as scientists are more likely to participate in scientific research discussions. The first reason for the result
is that this is in line with the characteristics of social media users’ occupations: freelancers, students, corporate employees, and public
institution employees account for more than 95%, that is, the number of general public on Twitter is much larger than that of other
occupational users. The second reason is that during the special period of pandemics, the scientific research results related to the
pandemic are also closely related to everyone, and some of them can solve various problems faced by people during the pandemic.
Especially considering the academic authority of high- cited scientific papers, general practitioners are also willing to spread more credible
pandemic-related information on Twitter (such as the high-cited pandemic papers on the subject of this study). Most of them hope that
their sharing behavior can help more people.

Table 5
Occupational types of Twitter mentions (total mentions = 425,754)
Occupational types No. of mention Percentage Average mention

counts counts

Members of the public 340942 80.1% 243.8783977

Scientists 49826 11.7% 35.64091559

Practitioners (doctors, other healthcare professionals) 26066 6.1% 18.64520744

Science communicators (journalists, bloggers, editors) 8670 2.0% 6.201716738

4.5.2 Geographical Distribution of Twitter Mentioned Users

Except for some users who did not indicate the national information data on Twitter (the invalid data of this study), the national
information of the users who participated in the mentioning of the global pandemic high-cited papers on Twitter are all counted, and the
valid data volume N = 170044. Figure 4 visualizes the regions mentioned by the paper on Twitter on the world map. The darker the color,
the more times users in that country are mentioned on Twitter. The results show that the Twitter users participating in the mention are
distributed in all continents except Antarctica (Fig. 5), of which North America represented by the United States and Canada occupies the
largest proportion, about 43.7%, followed by Europe 35.0%, Asia 10.7%, South America 6.9%, Oceania 3.3%, the lowest proportion is Africa,
only 0.4%. Table 6 counts the top 10 countries where papers are mentioned on Twitter. The results show that the United States is the
country with the most mentions by users, with 57,158 mentions, accounting for nearly 33.6% of total mentions, followed by the United
Kingdom with 35,838 mentions, accounting for nearly 15.2% of the total mentions, these two countries account for nearly 50% of all
country mentions. It can be seen that most of the users participating in the discussion of the pandemic’s high-cited papers on Twitter are
from English-speaking countries. After research and analysis, there are three reasons for this geographical distribution. First, Twitter users
are mainly concentrated in North America such as the United States, Canada, EU countries, and other developed countries such as Japan.
These countries have a large Twitter user base.

Page 9/19
 Table 6
Top 10 countries with papers mentioned on Twitter
Ranking Country Time of being mentioned Percentage(N = 170044)

1 America 57158 33.6%

2 Britain 25892 15.2%

3 Spain 15590 9.17%

4 Japan 11400 6.7%

5 Canada 9230 5.4%

6 Mexico 7855 4.6%

7 France 6738 4.0%

8 Australia 5555 3.3%

9 India 4889 2.9%

10 Brazil 4814 2.8%

Therefore, these countries are more on Twitter to discuss and exchange information about the high-cited papers of pandemics, and the
main reason for the low number of user mentions in China on Twitter is that the Chinese mainland has banned Twitter, so in China The
main software for dissemination of pandemic papers is WeChat and Sina Weibo. Second, most of the languages used in the papers in the
WoS are English. As the most common language in the world, English papers on pandemics are also easy to disseminate science in
English-speaking countries. Third, developed countries in European and American have formed a relatively good habit of disseminating
scientific information on social media. They are willing to use social media to communicate and share their own opinions or
others’research achievements on pandemics, and hope that with the help of the fast dissemination of social media, the public can better
understand the characteristics of pandemics, so as to help more people understand and fight against pandemics.
4.6 Feature Analysis of Mendeley Indicator
4.6.1 Mendeley readers category classification
Mendeley, as an online reference management tool, not only has the highest coverage among all Altmetrics indicators, but also has a
strong correlation with the citation frequency. It allows researchers to input statistics related to academic or professional status, position,
discipline and country (Edakar et al,2021). Table 7 shows the categories of readers of the high-cited pandemic papers on Mendeley.
Unidentified reader datas are not included in the statistics, and the remaining total number of readers is N = 648316. In this study, Mendeley
readers are classified as Doctor and Professor, Master and Postgraduate, Bachelor, Researcher, and others(Such as those who have not
attended college are others). Statistics show that the master and postgraduate readers occupies the largest proportion excepting 'others',
accounting for about 20.4%. Doctor and professor readers are also the main readers of the high-cited pandemic papers on Mendeley
platform, accounting for about 20.2% of the total readers, indicating that people with relatively high education are more likely to read
related pandemic literatures on Mendeley platform. In addition, the proportion of bachelor readers is 17.8% higher than that of researcher
readers 16.1%, which is contrary to people's common sense. Under the impression, most bachelors are engaged in basic education and
seldom engage in scientific research. However, the number of bachelor readers is higher than that of researchers, and is close to that of
doctoral and master readers. Indicating that the first is the advanced nature of bachelor education. the second is that the pandemic has
had a huge impact on people, and online platforms have increasingly become a tool for people to understand science and seek answers. In
addition, from the average number of readers per paper, it can be

Page 10/19
 Table 7
Mendeley reader category classification
Readers’ categories No. of reader Percentage(N = 648316) Average reader

counts counts

Doctor and Professor 130821 20.2% 93.6

Master and postgraduate 132202 20.4% 94.6

Bachelor 115318 17.8% 82.5

Researcher 104443 16.1% 74.7

Others 165592 25.5% 143.8

seen that the public's attention to the high-cited papers of pandemics is enthusiastic. However, reader identity categories are self-reported,
this may not reflect the true categories of readers, and some readers do not disclose their identity categories, and the real situation may be
somewhat different from the research in this paper.
4.6.2 Subject classification of Mendeley readers
The subject distribution of Mendeley readers of high-cited papers on pandemics is also a research objective of this paper. In this study,
readers who did not mark their disciplines on the Mendeley platform will be regarded as invalid data and will not be included in the
statistical data. The remaining valid reader data is N = 462907. Figure 6 shows the discipline classification of Mendeley readers of this
research paper, and Table 8 shows the top ten disciplines of Mendeley readers of this research paper. The data shows that the readers of
medicine and dentistry are significantly ahead of other disciplines, accounting for 36.7%. Because the research on global pandemics is

Table 8
Top 10 disciplines by Mendeley readers
Ranking Readers by discipline No. of reader Percentage(N = 462907)

counts

1 medicine and dentistry 169740 36.7%

2 Agricultural and Biological Sciences 57841 12.5%

3 Biochemistry, Genetics and Molecular Biology 57018 12.3%

4 Nursing and Health Professions 40540 8.8%

5 Social Sciences 31424 6.8%

6 Environmental Science 18937 4.1%

7 Immunology and Microbiology 18263 3.9%

8 Engineering 12996 2.8%

9 Psychology 12296 2.7%

10 Pharmacology, Toxicology and Pharmaceutical Science 11667 2.5%

closely related to the medical discipline, it is reasonable that the number of readers in this discipline is larger. The second and third majors
are Agricultural and Biological Sciences with 12.5% ​and biochemistry and genetics with

12.3%. The number of readers in these two disciplines is similar, indicating that readers of biology-related disciplines are also extremely
interested in research on global pandemics. The eight disciplines of design, chemical engineering, energy, physics and astronomy,
linguistics, materials science, mathematics, arts and humanities have less than 500 readers. This is consistent with previous studies,
which are low Mendeley readership for physics, mathematics, materials science. Surprisingly, the top ten reader’s disciplines are including
engineering and psychology. Generally speaking, these two disciplines, especially engineering, have little correlation with pandemics, but
this also reflects the global Pandemics, especially devastating pandemics like COVID-19, have caused great harm to the lives of all people,
and readers other than medicine-related disciplines, biology, and sociology are also concerned about pandemic research. Hoping to use the
power of science to fight pandemics.

Page 11/19
5. Discussion
In terms of measuring the influence of research papers, the academic community is increasingly calling for the need to combine the
trajectory of research achievements in online social media, exploring new ways to measure the impact of research results, so Altmetrics is
a good supplement to the citation frequency of papers. In the era of Internet of Everything and Big Data, the influence of scientific research
no longer depends on the number of citations, but on the number of people who share research on social platforms, media platforms, and
online academic platforms, participate in research and share research on social networks. Therefore, this paper aims to explore the
dissemination characteristic trajectories of the high-cited papers related to global pandemics included in the Web of Science database on
the online platform. Results show that 80.8% of the paper samples have been read, mentioned, shared or discussed on the online platform.
Due to social demands, researchers and scholars are increasingly being urged to communicate their findings to the general public(Regan et
al,2019). Especially during the quarantine period of pandemics, the need to share pandemic research results online is more urgent, which is
not only to serve as an early indicator of the impact of citations of pandemic papers, but also help people's lives. using Altmetrics
indicators can help identify important papers related to pandemics, and even guide people to read and share the important results of these
papers. in addition, It will also help the country understand the reading interest of scholars and the public during the pandemic, taking
effective measures to eliminate some tensions in society (Clarke L,2021). Although there is no broad consensus on the reliability of using
social network indicators to measure the impact of research papers, and some researchers are still reluctant or unaccustomed to share
their research results on the Internet. One of the main reasons is that the current Altmetrics indicators have not been fully formulated in the
evaluation system related to career development. But a series of statistical analyses based on data from a Swiss research center for
sustainable development, concluded that scientific research performance is generally positively correlated with online communication
effectiveness and participation in public promotion activities. This view has implications for academic incentive and evaluation
systems(Omar,2019).

Previous studies have confirmed that the coverage of Altmetrics indicators for research papers varies by research disciplines, nature, fields
and the type of web platforms used to collect Altmetrics indicators data (Hossseini E,2016). Our research findings show that research
results in the field of pandemics have better Altmetrics coverage, especially social media software represented by Twitter and Facebook,
and online reference management tools represented by Mendeley and Dimensions. Their indicator coverage rate is more than 70%. In the
years following the Covid-19 outbreak, due to the use of social media by scholars and researchers for research purposes has increased,
resulting in people are paying more attention to Altmetrics indicator data. And various studies have confirmed that many researchers prefer
to share their research results and academic contributions by taking advantage of the fast speed, wide range and high visibility of social
media to increase the popularity of their research (Sedighi et al,2020). The most-shared and most-discussed pandemic-related papers on
social media networks, if not the best, should be the ones that most resonate with the public. Our research also obtained the correlation
between the citation frequency, usage times of papers in Web of Science, AAS and some Altmetric indicators. We can obtain certain
Altmetric indicators that can be used as a supplement to the citation to measure the academic impact of the paper. Paper citations have a
strong correlation with the number of uses from 2013 to the present, and a moderate correlation with AAS. Secondly, a paper with a high
AAS means a high social influence, which conveys the current status of some scientific research or investigation to the public, and it is also
the scientific knowledge that the public urgently needs and pays attention to. That is to say, when users of various platforms in the network
conduct academic activities on the platform, they do not only look at whether the paper is cited a lot, whether the academic influence is
large, and other factors may be more dominant. Existing research confirms that Mendeley and Twitter, as tools for retrieving Altmetrics
data, are both strong indicators for research on social networks (Zahedi et al,2014;Erfanmanesh,2016;Delli et al,2017).

Notably, our study concludes that while Mendeley and Twitter both have high coverage and positively correlated with citations for high-
cited papers on pandemics. Mendeley is strongly correlated with citations while Twitter’s correlation is weak, which is enough to reflect the
different influences brought by the two platforms in the process of disseminating high-cited pandemic papers. In addition, this study also
draws information about the dissemination of research results of high-cited pandemic papers on other online platforms, such as
Facebook, Dimensions, News, Blogs, Policy sources, Weibo, CiteULike, Google+, Wikipedia and Reddit. Among them, the coverage rates of
Google+, CiteULike, Patents, Weibo, Redditors, and Vedio are all lower than 20%, indicating that the high-cited pandemic papers are less
disseminated and discussed in these social networks. Dimensions have a strong correlation with citation frequency, and the correlation
coefficient exceeds 0.95, which can reflect the academic influence of a paper almost as well as citation frequency. Facebook, blogs, and
policy are moderately related to citation frequency.

The main occupational types of users who mentioned high-cited pandemic papers on Twitter are members of the public, scientists,
Practitioners, and science communicators. The proportion of general public is much higher than that of other types of users. It can be seen
that general public actively disseminate pandemic-related information on Twitter (such as the high-cited pandemic papers). In terms of
geographical distribution of Twitter users, North America accounted for the largest proportion, followed by Europe, Asia, South America,

Page 12/19
Oceania and Africa. Most of the users come from English-speaking developed countries, followed by Japanese-speaking Japan and
French-speaking France. Because Twitter is banned in the Chinese mainland, China's Twitter mentions are relatively small. Research
analyzes that developed countries in Europe and the United States have formed a better habit of disseminating scientific information on
social media.

This research found that Mendeley readers with other categories account for about 20%. The first reason is related to the current COVID-19,
and the second is the universality of online document management platforms today. In addition, the number of bachelor readers is close to
that of doctoral and master readers, which reflects the advanced nature of today's bachelor education. In addition, senior researchers do
not have as many readers as other categories. Because Mendeley is a new literature management tool, senior researchers are not
accustomed to most social networking services and may choose to continue their traditional practice(Ahmed et al,2017). The readership
disciplines are dominated by medical-related majors, with more than 52%, because the research on global pandemics is most closely
related to the medical profession. In addition, biology-related majors, sociology, psychology, and engineering also have more readers. While
design, energy, physics and astronomy, linguistics, materials science, mathematics, arts and humanities readers are less than 500, which
also verifies that the number of Mendeley readers in the previous study of medical and social sciences is higher than that of physics,
mathematics and other disciplines.

6. Conclusion
Combining the existing research and our research results, tracking the dissemination trajectories and characteristics of scientific research
papers on social network platforms is conducive to a more comprehensive description of the impact of scientific research papers and the
value of scholars. Compared with the traditional paper measurement indicators, the new Altmetrics indicator can not only reflect the
academic impact of the pandemic high-cited papers, but also its social impact. In addition, some Altmetrics indicators are highly correlated
with traditional citation indicators, and can be used as supplementary scientometric indicators for evaluating the results of scientific
research papers and calculating the impact of research. The advantages of social media platforms are their ease of dissemination, wide
coverage, dynamic real-time data updates, and rich data on user mentions, engagement in discussions, and reader readings. The results of
this study show that Mendeley is the most important literature management platform for the dissemination and use of high-cited papers in
pandemics. And Twitter is the social media platform that is the most used resource in Altmetrics analysis. Promoting your own research
results and disseminating research results you think are important on Twitter is expected to receive good responses and views, which is
very helpful during the special period of the pandemic. In addition, the effect of using social media to disseminate scientific research
papers varies in different regions. For regions that do not do well in disseminating scientific research papers through social media and
literature management platform, the best approach is to educate and raise awareness of the dissemination of scientific papers and
research results in the network, of course, this requires a process. The research results show that the popularity of social network media
has brought convenience to most scholars and researchers in the pandemic environment, and it has also brought opportunities for
ordinary people in non-academic circles to understand scientific research. It's not so out of reach, the effect of this kind of communication
may be amazing, and the help to people is immeasurable. In addition, the government and universities should also consider gradually
incorporating Altmetrics indicators into the performance evaluation of teachers, scholars and researchers, which will help to further
promote their dissemination and sharing of scientific papers and research results on the Internet. Especially the research results related to
the pandemic, not only can it lead people to clearly recognize and understand the pandemic, but it can also remove some of the tension in
society.

There are also some limitations in this study. Firstly, the research papers are only obtained from a literature database, Web of Science. Only
two keywords and high-cited papers are used to search, and Altmetrics data acquisition only selected Altmetric.com as a data service
provider. The data sample may not necessarily reflect the transmission effect of overall global pandemics on network platforms. Secondly,
China is the country that publishes the second most international papers, this study did not count the data on the dissemination of papers
on WeChat and Tencent QQ, the two major Chinese social platforms, so it is impossible to know the dissemination effects and group
characteristics of pandemic papers by Chinese social network platforms. Thirdly, the information filled by users of various platforms may
be false, and the mention and reading of the paper may also have fraudulent data volume. This paper does not propose a solution to this
situation. Moreover, there is no continued exploration of the role of social media dissemination of scientific research papers in curbing the
spread of disinformation during the pandemic.

Finally, although AAS scores and various Altmetrics indicators can be used to supplement the impact evaluation of research results. Like
the existing literature and this study point out, Altmetrics does not indicate the quality of academic outputs, therefore, Altmetrics indicators
should be compared with traditional indicators, which achieve a more comprehensive and deep understanding of a paper's impact. Even
though the use of social media indicators for scientific evaluation are still quite a few problems to be solved. We should not lose sight of

Page 13/19
the fact that many network platforms are already widely used by scholars and researchers. Therefore, future research needs to further
explore the effect of these social networks on the dissemination of pandemic papers and the specific positive impact on society during the
pandemic, or continue to explore the role of social media dissemination of scientific research papers in containing pandemic
disinformation.

Declarations
Acknowledgments：

Funding: This work was supported by Science and Technology Commission of Shanghai Municipality Foundation (Grants No.
21511102904).

References
1. Chong M,Han W P . COVID-19 in the Twitterverse, from epidemic to pandemic: information sharing behavior and Twitter as an
information carrier[J]. Scientometrics,2021(1). https://doi.org/10.1007/s11192-021-04054-2.
2. Jiapei Li et al. Social media communication of the scientific and technological literature in emergency under COVID-19[J]. Library Hi
Tech,2021,39(3):796-813. https://doi.org/10.1108/LHT-10-2020-0268.
3. Rehemtula S , Rosario Arquero Avilés, Rosa M D L , et al. Altmetrics in Institutional Repositories: New Perspectives for Assessing
Research Impact[J]. Libraries in the Digital Age Proceedings, 2014.
4. Edakar M A M，Shehata A M K . Measuring the impact of COVID-19 papers on the social web: an altmetric study[J]. Global Knowledge
Memory and Communication, 2021, ahead-of-print (ahead-of-print). https://doi.org/ 10.1108/GKMC-11-2020-0179.
5. Ortiz-Núñez, Roelvis. Análisis métrico de la producción científica sobre COVID-19 en SCOPUS [J]. Revista Cubana de Información en
Ciencias de la Salud,2020,31(2):e1587.
6. Zhao Wanxin. Research on Altmetrics Indicators of Highly Cited Chinese Papers Based on Citation View [J]. Information
Studies:Theory & Application,2020,43(11):47-53. https://doi.org/10.16353/j.cnki.1000-7490.2020. 11. 008.
7. Priem J, Taraborelli D, Groth P and Neylon C. (2010),“Altmetrics:A Manifesto”, available at:http: // altmetrics.org/manifesto/.
https://doi.org/10.1234/12345678.
8. Yao Li.Research on the Altmetrics Evaluation System of Academic Books’Impact in China[D].Nanjing University,2017.
https://doi.org/10.27235/d.cnki.gnjiu.2017.000394.
9. LaneJ.Let’s make science metrics more scientific[J].Nature,2010,464(7288):488－489. https://doi.org/10.1038/ 464488a.
10. Wu Shengnan.Study on the Method and Application of Altmetrics[D].Wuhan University,2015.
11. Finbar Galligan,Sharon Dyas-Correia. Altmetrics: Rethinking the Way We Measure[J]. Serials Review, 2013, 39(1) : 56-61.
https://doi.org/10.1016/j.serrev.2013.01.003.
12. Dinsmore Adam,Allen Liz,Dolby Kevin. Alternative perspectives on impact: the potential of ALMs and altmetrics to inform funders
about research impact [J]. PLoS biology, 2014, 12(11) : e1002003. https://doi.org/10. 1371 /journal.pbio.1002003.
13. Yang Liu,Chen Ming. A Comparative Study of Commonly Used Alternative Metrology Tools [J]. Information Studies:Theory &
Application,2015,38(09):114-119+144. https://doi.org/10.16353/j.cnki.1000-7490.2015.09.023
14. Wang Xiuhan. Research on paper influence evaluation from the perspective of Altmetrics [D]. Tianjin Normal University,2021.
https://doi.org/10.27363/d.cnki.gtsfu.2021.000173.
15. Hammarfelt B. Using Altmetrics for assessing research impact in the humanities[J].S- cientometrics , 2014, 101 (2):1419-1430.
https://doi.org/10.1007/s11192-014-1261-3.
16. Banshal SK,Singh VK. Disciplinary Variations in Altmetric Coverage of Scholarly Articles [C]//17th Intertional Conference On
Scientometrics& Informetrics. 2019. https://doi.org/10.48550/arXiv.1910.04205.
17. Alperin J P. Geographic variation in social media metrics:an analysis of Latin Amer-ican journal articles[J]. Aslib Journal of
Information Management,2015,67(3):289-304. https://doi.org/10.1108/ajim-12-2014-0176.
18. Robinson Garcia N, Torres-saunas D, Zahedi Z, et al. 2014. New data, new possibil-ities: exploring the insides of Altmetric.com [J]. EI
Professional De La Informacion,23(4):359-366. https://doi.org/10.3145/epi.2014.jul.03.
19. BALKE W T.What does Twitter measure?: influence of diverse user groups in Altmetrics [C].ACM/IEEE-Cs Joint Conference on Digital
Libraries,2015:119-128. https://doi.org/10.1145/2756406.2756913.

Page 14/19
20. Huang Xiao,Gao Jiahui,Wu Jiang. Analysis on Altmetrics Indicators Characteristics of Highly Cited Papers in Different Disciplines [J].
Information Studies:Theory & Application,2019,42(09):56-63. https://doi.org/ 10.16353/ j.cnki.1000-7490.2019.09.010.
21. Fang Zhichao,Wang Xianwen. Study on the Accumulation Speed and User Type of Scientific Publications’ First Tweets [J].
Documentation,Information & Knowledge, 2019(02): 28-38. https://doi.org/10.13366/j.dik.2019. 02.028.
22. Costas R,Zahedi Z,Wouters P . Do "altmetrics" correlate with citations? Extensive comparison of altmetric indicators with citations
from a multidisciplinary perspective[J]. Journal of the Association for Information Science and Technology, 2015, 66(10).
https://doi.org/10.1002/asi.23309.
23. Thelwall M,Haustein S,V Lariviére,et al. Do altmetrics work?: Twitter and ten other candidates[J]. PloS One, 2013, 8(5): e64841.
https://doi.org/10.1371/journal.pone.0064841.
24. Mohammadi E,Gregory K B,Thelwall M,et al. Which health and biomedical topics generate the most Facebook interest and the
strongest citation relationships?[J]. Information Processing & Management, 2020, 57(3):102230.1-102230.
https://doi.org/10.1016/J.IPM.2020.102230.
25. Haustein S,Peters I,Sugimoto C R,et al. Tweeting biomedicine: an analysis of tweets and citations in the biomedical literature[J].
Journal of the American Society for Information Science & Technology, 2013, 65(4):656-669. https://doi.org/10.1002/asi.23101.
26. Jiapei Li,Seong, et al. Construction of Scientific Impact Evaluation Model Based on Altmetrics[J]. Journal of information and
communication convergence engineering, 2017, 15 (3): 165-169. https://doi.org/10.6109/jicce. 2017.15.3.165.
27. Chen Huafang,Xiang Fei. The Characteristics and Advantages of Altmetrics in Academic Papers Evaluation: A Case Study of Medical
and Health Sciences [J]. Digital Library Forum,2021(02):19-26.
28. You Qingbin,Wei Bo,Tang Shanhong. Evaluation Model Construction to Evaluate Article’s Influence Based on Altmetrics [J]. Library and
Information Service,2014,58(22):5-11. https://doi.org/10.13266/j.issn.0252-3116. 2014. 22.001.
29. Vysakh C,Babu D H R . An Altmetric Approach to Measure the Social Media Attention of COVID-19 Articles[J]. Library Philosophy and
Practice, 2020.
30. Chong M,Han W P.COVID-19 in the Twitterverse, from epidemic to pandemic: information-sharing behavior and Twitter as an
information carrier[J]. Scientometrics, 2021(1). https://doi.org/10.1007/s11192-021-04054-2.
31. Moon J Y,Yoon D Y,Hong J H,et al. The Most Widely Disseminated COVID-19-Related Scientific Publications in Online Media: A
Bibliometric Analysis of the Top 100 Articles with the Highest Altmetric Attention Scores[J]. Healthcare, 2021, 9(2):239.
https://doi.org/10.3390/healthcare9020239.
32. Fabiano N , Hallgrimson Z , Kazi S , et al. An analysis of COVID-19 article dissemination by Twitter compared to citation rates. Cold
Spring Harbor Laboratory Press, 2020. https://doi.org/10.1101/2020.06.22.20137505.
33. Tornberg H N,Moezinia C,Wei C,et al. Assessment of the dissemination of COVID-19 articles across social media-An Altmetrics and
PlumX Study (Preprint).2020. https://doi.org/10.2196/preprints.21408.
34. Roy D, Das M, Deshbandhu A. Postcolonial pandemic publics: examining social media health promotion in India during the COVID-19
crisis [J]. Health Promotion International,2021. https://doi.org/10.1093/heapro/daab076.
35. Wu Shengnan,Zhao Rongying. Research on Status and Development Trend of Altmetrics Tools [J]. Documentation,Information &
Knowledge,2016(01):84-93. https://doi.org/10.13366/j.dik.2016.01.084.
36. Khosla I . Book Review: Social Research Methods: Qualitative and Quantitative Approaches[J]. Frontiers in Psychology, 2021,
12:696828. https://doi.org/10.3389/fpsyg.2021.696828.
37. Shi Li,Qin Ping,Li Xiaotao. Analysis on Relationship between Interdisciplinarity and Altmetrics Indicators of Highly Cited Papers [J].
Information Studies:Theory & Application,2021,44(05):60-65+91. https://doi.org/10.163 53/j.cnki.1000-7490.2021.05.010.
38. Ou Guiyan,Ma Yanting. Analysis of the Characteristic Factors and the Evolution of the High Altmetrics Score Papers in 2013-2018 [J].
Journal of the China Society for Scientific and Technical Information,2020,39(03):243-252. https://doi.org/CNKI:SUN:QBXB.0.2020-03-
002.
39. Li Xiaotao,Jin Xinyi. Research Hotspots and Frontier Analysis of Scientometrics Based on Altmetrics［J］.Journal of Modern
Information,2019,39(1):153-160. https://doi.org/10.3969/j.issn.1008-0821.2019.01.019.
40. Tan Rongbo,Mei Xiaoren. SPSS statistical analysis practical tutorial [M]. Beijing: Science Press,2007.
41. Regan I,Henchion M.Making sense of altmetrics: The perceived threats and opportunities for academic identity[J]. Science and Public
Policy, 2019, 46(4):479-489. https://doi.org/10.1093/scipol/scz001.
42. Clarke L . Covid-19: Who fact checks health and science on Facebook?[J]. BMJ, 2021, 373:n26-. https:// doi.org/10.1136/bmj.n1170.

Page 15/19
43. Omar K . Does public outreach impede research performance? Exploring the ‘researcher’s dilemma’ in a sustainability research
center[J]. Science and Public Policy, 2019,46(5):420-420. https://doi.org/10.1093/scipol/ scz024.
44. Hossseini E.Altmetrics for Information Professionals: Past, Present and Future[J].INFORMATION RESEARCH-AN INTERNATIONAL
ELECTRONIC JOURNAL, 2016,21(4).
45. Sedighi M.Evaluating the impact of research using the altmetrics approach (case study: the field of scientometrics)[J]. Global
Knowledge Memory and Communication, 2020, ahead-of-print(ahead-of-print). https:// doi.org/10.1108/GKMC-02-2019-0013.
46. Zahedi Z,Costas R,Wouters P.How well developed are altmetrics? A cross-disciplinary analysis of the presence of 'alternative metrics'
in scientific publications[J]. Scientometrics, 2014, 101(2):1491-1513. https://doi.org/10.10 07/s11192-014-1264-0.
47. Erfanmanesh M , Abdullah A , Karim N . Information Seeking Anxiety: Concept, Measurement and Preliminary Research[J].
International Journal of Information Science and Management, 2014, 12(1):47-64.
48. K, Delli, C, et al. Measuring the social impact of dental research: An insight into the most influential articles on the Web.[J]. Oral
Diseases,2017. https://doi.org/2017. 10.1111/odi.12714.
49. Ahmed, Shehata, David, et al. Changing styles of informal academic communication in the age of the web[J]. Journal of
Documentation, 2017, 73(5). https://doi.org/10.1108/JD-06-2016-0083.

Figures

Figure 1

Example of AAS for the Altmetric.com

Page 16/19
Figure 2

Topics Clustering Map of Highly Cited Pandemic Papers

Figure 3

AAS Score Ranking (SR) distribution

Page 17/19
Figure 4

Geographical distribution of Twitter mentions

Figure 5

Paper mentions on Twitter by continent

Page 18/19
Figure 6

Subject distribution of Mendeley readers

Page 19/19