Page 177 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 42

176 Journal of Library Science in China, Vol. 8, 2016

some certain fields, which may function for preference, retrospection or comparison. These kind of
sharing and usage cannot be captured by citations.
(4) Weibo altmetrics indicator was significantly concentrated and scattered. In the dataset, each
scholarly article had on average 3.46 Weibos, but 65.48% of the articles had only 1 Weibo. And
percentage of articles that received less than 3 Weibos reached 83.84%, while percentage of articles
that received more than 10 weibos was only 5.86%. This indicated most Weiboed article received
relatively low Weibo attention. On the other hand, 1% of the Weiboed articles obtained 30% of all
Weibo attention, 5.1% of the Weiboed articles obtained 50% of all Weibo attention, and 33.4% of
the Weiboed articles obtained 80% of all Weibo attention. This demonstrated that small percentage
of articles were extremely highly Weiboed.
(5) Altmetrics score of Weiboed articles were much higher than the average level. This remained
true when taking journal and time factors into consideration. Altmetrics score of Weiboed articles
were 23.24 times (higher than 86.98%) over that of articles in the whole dataset, and 15.42 times
(higher than 86.08%) over that of articles in the whole dataset in latest 3 months. In addition,
altmetrics score of Weiboed articles were on average 19.41 times over (higher than 82.39%) that
of articles in the same journal, and 22.36 times over (higher than 70.86%) that of articles in the
same journal in latest 3 months. Except for few cases compared with article in the same journal
in latest 3 months, Weiboed articles all had altmetrics score higher than the average level. These
demonstrated that article mentioned and discussed by Weibo were basically hot international
scholarly articles, and Weibo altmetrics indicator could be potentially used to predict the worldwide
attention.
Weibo, as the biggest microblog service in China, has been increasingly used by scholarly
institutions and individual scholars to follow, disseminate and discuss scholarly artefacts, which
conveys scholarly value and social value beyond traditional citations, and is an important
altmetrics indicator data source. Many other academic tools and services, for example, ScineceNet.
cn, Xinkexue.com, datatang.com, Wanfang academia etc. all record scholarly traces and could
be mined, used and applied eventually to science and society. While mining the value of these
records and data, scholars should think out of the box and make full use of modern online tools to
do scholarly communication, for example, microblog can be used in conference communication,
propagation and education and research results recommendation etc., in order to improve the
efficiency of scientific work and usage of scientific results.
The paper studied Weibo altmetrics indicator, but would provide reference for studying other
altmetrics indicator. The international microblog service Twitter has been relatively thoroughly
studied. Weibo has different cultural background and language features from Twitter. It requires
further comparative analysis to see whether they have different characteristics. For example,
with the same limit of 140 characters, Chinese is able to express richer content than English does
(Zhang & Pentina, 2012). As altmetrics studies dive deeper, altmetrics researchers have realized
that counts of altmetrics indicators make sense only when combined with content. This is both