EFFECTIVENESS OF SOCIAL NETWORKING
IN A UNIVERSITY LIBRARY ENVIRONMENT
Hou Ieong Ho and Jieh Hsiang
Department of Computer Science and Information Engineering, National Taiwan University, Taipei, Taiwan
Keywords: Social network, Library users, Quantitative analysis, Library services.
Abstract: As social networking becoming commonplace among the young patrons of the libraries, many libraries are
incorporating their services into existing social networking environments. However, there have not been
many studies on the effectiveness of such services. In this paper, we describe how the National University
Library utilizes two social networking platforms, Facebook and Plurk, to provide better services to its
webized community. The effectiveness of these services is demonstrated through a quantitative analysis.
Different behaviour due to the different propagation mechanisms of Facebook and Plurk is also discussed.
Our results indicate that certain library services using social networking services are easy to build, easy to
maintain, easy to analyze, and are effective for reaching out to webized patrons.
1 INTRODUCTION
A university library serves a central role as a
provider of academic material and information
services to its patrons: faculty, researchers and
students of the university. This role, however, has
been threatened by the rapid growth of the World
Wide Web with its abundant information and easy-
to-use services. Young patrons, especially
undergraduate and graduate students, not only use
the Web as a major source of information, but also
as their personal working and social environments,
among the most popular are Facebook, MySpace,
Twitter, and Plurk. We call these patrons “webized
patrons”, whose daily life and social activities are
often closely tied with the Web through social
networking. Social networking platforms, thus,
should also provide the libraries a new channel to
reach out to their webized patrons. Indeed, such an
approach has been advocated (Breeding 2007;
Breeding 2009; Chu & Meulemans 2008; Harris &
Lessick 2007; Milstein 2009) and some libraries
have already set up such services. However, we have
found little discussion or analysis of the
effectiveness of these library services. In this paper
we describe the social networking services provided
by the National Taiwan University Library
(NTULIB). We also provide a quantitative analysis
of the effectiveness of these services. Our studies
show that they are easy to build and effective in
practice.
The services we describe are provided via
Facebook and Plurk, two of the most popular social
networking platforms among college students in
Taiwan. They also use different mechanisms to
propagate information, which makes our
experiments even more interesting. There are three
phases in our experiments. We first incorporated a
social bookmarking service to provide evidence that
there are indeed webized patrons in the community
that the library intends to serve. In the second phase
we incorporated different services of NTULIB into
Facebook and Plurk without announcing them
officially. In the case of Facebook, NTULIB
acquired 2,400 fans, 90% of which under the age of
34. This number is quite significant for a university
with a student population of about 30,000. After
announcing the service (in the third phase) officially,
the number jumped to 3,572 in 3 months and 5,686
in 8 months. Plurk’s way of defining friends and
fans is less transitive than Facebook and thus makes
the rippling effect less transparent. However, the
NTULIB Plurk page still acquired 150 friends before
the official announcement and 644 afterwards. The
number of friends reached 1,057 after 8 months. In
another interesting experiment, we announced a
friends-only service (only friends of NTULIB Plurk
could see) within our webpac that required a
password to use. During that week, we saw an
307
Ieong Ho H. and Hsiang J..
EFFECTIVENESS OF SOCIAL NETWORKING IN A UNIVERSITY LIBRARY ENVIRONMENT .
DOI: 10.5220/0003079803070310
In Proceedings of the International Conference on Knowledge Management and Information Sharing (KMIS-2010), pages 307-310
ISBN: 978-989-8425-30-0
Copyright
c
 2010 SCITEPRESS (Science and Technology Publications, Lda.)

increase of 2,000 hits to our webpac. This shows the
word-of-mouth effect of Plurk.
We feel that our experiments demonstrate the
positive effect of social networking services to reach
out to webized patrons in a university library
environment. Incorporating these services is
inexpensive and all the statistical analysis tools and
logs are readily available from the Web. We hope
our studies will encourage more libraries to use
these services.
This paper is organized as follows. We first
present a simple study that shows that there are
indeed webized patrons at NTU. We then describe
our services and experimental results on the
Facebook platform, followed by the same on Plurk.
Further experiments that demonstrate the different
behavior caused by the different propagation
mechanisms in the two services are also given. The
paper concludes with a brief discussion.
2 SOCIAL BOOKMARKING IN
NTULIB
Before launching social networking services at
NTULIB, we ran a small experiment to find out if
there is indeed a webized population among our
patrons. This pre-project is necessary to be certain
that the webized community and the library patrons
within the university environment are not mutually
exclusive.
This experiment was done via a social
bookmarking service, launched in the summer of
2007. This is a “Bookmark and Share” function
(provided by addthis.com http://addthis.com) on the
detailed (catalogue) page of every book item that our
patrons find through the NTULIB webpac (see
Figure 1).
Figure 1: The social bookmarking service in NTULIB’s
webpac.
This feature allows the patron to either add the
book item to his browser’s favorite list, email the
item to someone, print out the page, or add the item
to one of a list of social bookmarking websites
(including Facebook, Twitter, MySpace, and others).
This service is added at this particular point in a
webpac session because it is the last web page that a
patron sees before checking out a book. Figure 2
shows the usage numbers between July 2007 and
April 2010 (obtained from logs provided by
addthis.com). Comparing to other options provided
in the function, social networking wins out every
month. Furthermore, the numbers of social
networking users never dropped below 200 except
once since December 2008. Even during winter
holiday, February 2010, the number was still more
than 200.
Figure 2: NTULIB webpac bookmark numbers.
This experiment reveals that there is indeed a
fairly significant number of webized patrons in our
library community.
Further reviewing the logs shows that the largest
percentage of those uses involves Facebook. Thus
we chose Facebook as one of the testbeds. We also
chose Plurk as the second testbed because (1) its
social networking mechanism is different from
Facebook, and (2) it is more popular among college
students in Taiwan than other similar services such
as Twitter (Lai 2009).
3 NTULIB IN FACEBOOK
We set up a Facebook fans page
(http://www.facebook.com/NTULIB) in April 2009.
In order to test how the ripple effect of social
networking works on its own, we intentionally
withheld the announcement of the availability of this
service. Our Facebook page contains library
information, librarian selected library events, library
news, and an NTULIB Search box. New messages
are posted irregularly. Six months after being
KMIS 2010 - International Conference on Knowledge Management and Information Sharing
308

registered with Facebook, our page acquired 2,400
fans (note, without any public announcement). The
way the propagation works is that whenever a user
becomes a fan of the NTULIB page, gives a “like”
flag to an NTULIB announcement, or leaves a
comment, it will be recorded on that user’s own
Facebook page. This information can be seen by the
user’s friends, who may choose to become a fan
himself; thus propagating the NTULIB page.
We then gave the first public announcement in
November 2009. By 14 July 2010, the number of
NTULIB fans increased to 5,686 (in comparison,
ALA Library has 3,236 fans at that time
(http://www.facebook.com/alalibrary)). After
analyzing the statistics provided by Facebook, we
observed that almost 90% of our Facebook patrons
are under the age of 34, and that there are more
undergraduate students than graduate students.
The NTU Search box, which incorporates the
NTULIB webpac interface in the Facebook
platform, however, produced mixed results. There
were only 36 active users of this service in the
month of January 2010 and 21 active users in July
2010. Such low returns were also observed in
services of other libraries (see Table 1). Indeed, even
global library services such as JSTOR and WorldCat
did not show impressive numbers. This seems to
indicate that social networking is still largely a
social activity and is somewhat disconnected with
the patrons’ academic activities. How to improve
this situation remains a challenge.
Table 1: Monthly active users of library applications in
Facebook, sampled in 23 Jan, 24 May and 19 July, 2010.
Application name
Monthly Active Users
In
App
Jan
2010
May
2010
July
2010
JSTOR Search
1,376 572 298 Y
WorldCat
765 924 809 Y
UM Library Search
242 17 9 Y
Penn State University
Libraries Search
47 28 27 Y
UQ Library Search
41 28 11 N
Loughborough University
Library
37 50 26 N
NTULIB Search
36 20 21 Y
Mississippi State
University Library App
22 24 12 N
4 NTULIB IN PLURK
Similar to Twitter, Plurk is a free micro-blogging
service that allows users to post micro-blogs (called
“plurks”) of up to 140 text characters in length.
There are, however, some differences between Plurk
and Twitter. Someone who follows a user is called a
fan, and two mutual fans are friends. A user’s Plurk
page shows the plurks written by the user, his
friends, and the people of whom he is a fan. They
are shown in chronological order using a timeline,
with replies to the same plurk grouped together as a
thread. However, Plurk does not show the change of
activities of friends (and fans) like is done in
Facebook. Therefore it is harder for a new Plurk to
be noticed.
We set up a Plurk account
(
http://www.plurk.com/NTULIB) on 15 July 2009. The
NTULIB Plurk provides library announcements,
new books suggestions from the NTULIB’s
Featured New Books Blog
(
http://newbooks.lib.ntu.edu.tw) and an integrated
webpac service. We wrote a program that transports
library announcements as new plurks every hour (if
any). A daily summary of new book suggestions is
also provided automatically.
Since Plurk does not provide ways to incorporate
plug-in APIs, we provide NTULIB webpac search
facility in a different way. We implemented a
software agent that automatically searches for
keywords such as “find book” in any new plurks of
friends of NTULIB. The program will then use the
title of the book to perform a search in the NTU
webpac and, if finds books that match, returns the
links to the user.
Similar to the experiments with Facebook, we
did not make an announcement of the NTULIB
Plurk when the service was launched in July 2009. It
acquired 150 friends and 30 fans before the official
announcement was made in November 2009. By 14
July 2010, it has accumulated 1,057 friends and 145
fans.
The propagation mechanism of Plurk is more
limited than Facebook. Plurk does not provide
updates on the activities of friends. Neither does
Plurk provide facilities that cite other plurks (such as
“retweet” in Twitter). Thus our Plurk page is
propagated only through friends’ plurks and auto-
replied messages. Patrons may also write plurks to
tell their friends about NTULIB plurk page out of
their freewill. This lack of transitivity makes the
accumulation of friends and fans more challenging.
By analyzing the statistics provided by Google
Analytics between 1 November 2009 and 17 July
EFFECTIVENESS OF SOCIAL NETWORKING IN A UNIVERSITY LIBRARY ENVIRONMENT
309

2010, we noted that there were 1,448 visits (via 656
different Plurk pages) from Plurk to the NTULIB
webpac system. Plurk also ranked 1st in visits (from
external sites) to the NTULIB’s Featured New
Books Blog, recording 2,024 such visits (Figure 3).
Figure 3: Top 10 referral domains to NTULIB Featured
New Books Blog.
Because of Plurk’s lack of transitive
propagation, we made another experiment to see
how information propagates through Plurk. On the
early morning of 24 December 2009, we made a
friends-only announcement (to only the 644 friends
of NTULIB Plurk) that there was a limited-time only
hidden Christmas greeting hidden in our webpac that
can only be accessed with a password. When
analyzing the logs later, we noted a peak with 2,000
more visits to our webpac during that period. This
was quite surprising exactly because of the lack of
transitive propagation in Plurk. In other words, these
extra visits we received can only be caused by the
word-of-mouth rippling effect of Plurk users.
5 DISCUSSION
In this paper we showed that social network can be
an effective new medium for reaching out to
webized patrons. We demonstrated this point by
conducting experiments using two popular social
networking platforms, Facebook and Plurk. We first
introduced a social bookmarking service to justify
that there are indeed webized patrons in our
university environment. We then showed the
effectiveness of our services first through
unannounced services, and then showed the
differences by officially announcing them eight
months later.
Our analysis shows that there is indeed a
webized population among university librarian
patrons. Even when the social networking services
were not announced, they already attracted a
significant size of users. However, the relatively
small numbers of users that use the webpac search
services also indicate that these patrons use our
services mainly to keep up to date on the various
activities such as exhibitions and new book
announcements that they library provides.
Another important feature is that the services
that we described are technically easy to implement
and require very little additional resources to
maintain. Even the analysis tools and traffic logs are
readily available and free. This is particular useful
since library budgets everywhere are shrinking. (The
necessity of additional human resources as required
in (Milstein 2009) is also why we do not recommend
that approach.) Using traffic logs and statistics from
the Web should also be more accurate in capturing
the user behavior of webized patrons than traditional
surveys since the data are implicitly extracted from
their working environment.
The difference in the ways of propagating
information in Facebook and Plurk is also worth
mentioning. Since Facebook is more transitive in the
propagation of activities of friends, it is easier to
accrue a large number of fans. On the other hand,
having lots of fans does not mean that they are all
interested in library-related activities. In comparison,
the relative lack of transitivity in Plurk could
produce more loyal followers, as indicated through
the analysis of the logs and our last experiment with
the Christmas greetings.
REFERENCES
Breeding, M. 2007, ‘Librarians face online social
networks’, Computers in Libraries, vol. 27, pp. 30-33.
Breeding, M. 2009. ‘Social Networking Strategies for
Professionals’, Computers in Libraries, vol. 29, pp.
29-31.
Chu, M. & Meulemans, Y. N. 2008, ‘The Problems and
Potential of MySpace and Facebook Usage in
Academic Libraries’, Internet Reference Services
Quarterly, vol. 13, pp. 69-85.
Harris, A. & Lessick, S. 2007, ‘Libraries Get Personal:
Facebook Applications, Google Gadgets, and
MySpace Profiles’, Library Hi Tech News, vol. 24, pp.
30-32.
Lai, H.-C. 2009, ‘Exploiting Cloud Computing for Social
Network Analysis － Exemplified in Plurk Network
Analysis’, TAAI, Taiwan.
Milstein, S. 2009, ‘Twitter for Libraries and Librarians’,
Computers in Libraries, vol. 29, pp. 17-18.
KMIS 2010 - International Conference on Knowledge Management and Information Sharing
310