Collective Intelligence
How Collaborative Contents and Social Media Changing the Face of Digital
Library
Agnes Devina Haryuni, Yong Zhang and Chunxiao Xing
RIIT, TNLIST, Department of Computer Science and Technology, Tsinghua University, Beijing, China
Keywords: SNS, Social Network, Digital Library, Collective Intelligence, Collaborative System, Social Networking
System, Social Media, Recommender System.
Abstract: The growth of digital libraries has provided useful resources for users. However most digital libraries are
not effectively to promote themself and engage audience. Limited resources also postponed the growth of
the materials collection and expansion of related research and discussion. Our study shows how digital
library can use social networking system to promote new material and engage user, allow user to share
information and thoughts. It facilitates a collective intelligence system where user can interact and
contribute in a knowledge sharing environment. In a platform for user-generated content, user can submit
material and join discussion, and admin will learn the inputs and filter the content. This paper proposes the
use of social networking system in digital library, delivers some case studies and explains how we can use
social media to expand user base with recommender method. In this paper we also provide
recommendations for further development.
1 INTRODUCTION
Many virtual intermediaries have been used to
digitalized online archives, provided research-
based file-sharing, and facilitated collective
knowledge sharing in discussion forum.
Levy and
Marshall (1995) regarded that digital libraries are
collections containing fixed, permanent documents
which are based on digital technologies and are
used by individuals working alone. We aim to
improve the current involvement of collective
knowledge sharing and social networking system
for digital library in a context of open
knowledgebase platform.
The current user involvement in digital libraries
is still very low. Most institutions only provide
static format of their digital library with little or no
promotion to the students and visitors. Users find it
hard to search for certain issue and since they
rarely learn about digital library, there are very few
people who are interested to explore it by
themselves. Until now, digital library is still
considered as a time-spending research resources
with limited advantages.
We aim to import social network system to
digital library, where users can easily find digital
content, we can easily preserve the information and
share it, and users will also be more involved in
generating content, sending comments and
feedback and exploring others’ opinion about
certain data. This will be a crowdsourcing method,
where people from different locations can
contribute together toward a cause for the
community.
Boyd and Ellison (2007) mentioned that social
network sites are web-based services that allow
individuals to construct a public or semi-public
profile within a bounded system, articulate a list of
other users with whom they share a connection,
and view and traverse their list of connections and
those made by others within the system. In effect,
social recommenders can leverage users’
acquaintance with the recommendation source,
which instantly attaches a wealth of established
social information to the recommendations that can
be further explored and exploited in the processes
of inspection and control (Groh et al., 2012).
This means it’s necessary to build a bounded
connection, understand the needs, and know their
connection with others. Collective intelligence
characterizes multi-agent, distributed systems
349
Devina Haryuni A., Zhang Y. and Xing C..
Collective Intelligence - How Collaborative Contents and Social Media Changing the Face of Digital Library.
DOI: 10.5220/0004353403490354
In Proceedings of the 9th International Conference on Web Information Systems and Technologies (WEBIST-2013), pages 349-354
ISBN: 978-989-8565-54-9
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
where each agent is uniquely positioned, with
autonomy to contribute to a problem-solving
network (Gill, 2012).
Inside an e-learning scenario, the concept of
digital library system naturally translates into a
virtual environment, where interactions are
welcomed and eased, and where every community
service, like wikis and forums, contributes to the
creation of a common knowledge as part of a
structured learning process.
Digital library is one of the service for virtual
learning and an innovative system of collective
intelligence to make it easier for user to learn and
promote online learning environment. Di Cerbo,
Dodero, and Succi (2008) encouraged a virtual
environment to facilitates community service such
as e-learning service. In this paper, we find out
how social networking approach can expand user
base and encourage collective intelligence, which
if executed properly, it will result in a series of up-
to-date materials, new suggestions, amateur inputs
in discussion related to subjects and cutting cost
and effort to expand the collection while users
offer new contents.
An interactive virtual medium will also connect
different perspectives and resources in related
fields and provide additional knowledge base
necessary to expand research and discussion for
the topic in question. Collaborative systems are
tools used to facilitate the implementation of group
work (Aparicio and Costa, 2012).
In this paper, we investigate some case studies
about how social networking system deployment is
visible for various uses such as digital library
expansion and integrated collective knowledge
base. Regarding to certain weakness in the
application method so far, we plan an improved
system for future work and explain the detail in
architecture design. Finally we recommend some
necessary suggestions for future expansion.
This paper consist of: Section 2 introduces
collective intelligence use cases. Section 3 presents
details of social networking system process and
case studies of social networking involvement and
recommender system. Section 4 explores how to
combine social networking system with collective
intelligence concept. Section 5 has an
implementation plan, along with arcitecture design.
Section 6 explains the recommendation for future
work.
2 COLLECTIVE
INTELLIGENCE USE CASES
In software development, collective intelligence
offers a solution to complete a system or run
analysis and maintenance in a cost-effective and
time-effective way.
Collective intelligence approach has been used
in web-based application and software engineering,
where participants are engaged virtually to a global
project that combines each of their work result to
complete a database, or to construct a new system
or expand and improve an already established
system. Such projects can benefits a massive
choices of ideas, technical support and knowledge
base for some various participants, from amateur to
professionals.
2.1 Knowledge Sharing in Informal
Communication
The Tree of Knowledge (Kwon et al., 2011) is an
enabling technology that helps people share their
knowledge through their ongoing informal
interactions with their colleagues in a specific
place. It simulates the information-sharing
environment where the interaction is projected as
the tree withers and the leaves fall. The states of
the tree environment, such as sunshine, windy,
snowstorm, and other weather conditions, also
depends on the level of interaction. When people
approaches the tree, the client system will load the
welcome page.
The current prototype of the Tree of
Knowledge is a web-based distributed system
consisting of multiple clients, a data server, and the
tree system. The client side’s user interface
displays the postings and a textbox to post a
message. When the users post their knowledge
including ideas, questions, comments, critiques,
notes, and random thoughts, the location tag is
automatically attached and the user-defined tags
can be attached to the message.
2.2 Personalized Ebooks Learning
Application
Social network growing features keep on providing
users with new medias to interact and learn.
Ribière, Picault, and Squedin (2010) developed an
application to facilitate cross-media and cross-
community information discovery; facilitate
information discovery with contents of all sorts
WEBIST2013-9thInternationalConferenceonWebInformationSystemsandTechnologies
350
from all sources; extend the e-book concept to be a
dynamic collection of multimedia contents from all
sources and extend reading to discovery for formal,
leisure and spontaneous browsing and learning.
The system finds users of the social network of the
book that have annotations in common. Then it
analyzes sequences of annotations of those people,
compares them with the sequence of annotations
each user made, and at the end, suggests possible
next steps within the book, in terms of contacting
opportunities or reading opportunities.
3 RECOMMENDER SYSTEMS
IN SOCIAL NETWORKING
As an effective social networking method,
voluntary social campaign is also similar as
recommender system. Recommender system is a
system that suggests certain materials for a user in
the acknowledgement that the user in question has
read, explored, searched, and discussed about a
related topic. While one method uses real user and
another one uses an machine-based automate
approach, both manage to generate more users to
increase our user base, and in return, give a bigger
chance to collect more resources through user-
generated content uploads.
3.1 RecDB: Social Recommender
System
The main purpose of recommender system named
RecDB is to suggest users with useful and
interesting items or content (data) from a
considerably large set of items (Sarwat and
Mokbel, 2012). RecDB is an efficient and scalable
system that provides online recommendation to
users. Online recommendation viewed by users
comes from the system function inside the
database engine. The module, Rec-tree, a multi-
dimensional tree index structure that is built
specifically to index recommendation models and
provide flexible and online recommendation to
users. As in traditional database index structures,
the user can define which users/items attributes
(i.e., dimensions) are needed to be indexed by Rec-
tree.
3.2 Tasteweights Music
Recommendation Tool
Social recommendation spreading information to
virtual friends, and often the information details
depend on how close in the relationship between
users. Bostandjiev, O’Donovan, and Höllerer
(2012) studied this approach by developing
TasteWeights, a social recommender tool. The
TasteWeights system recommends new
artists/bands based on the music “likes” of the user
and her Facebook friends. TasteWeights system
displays a graph that shows the users’ items, their
friends, and the recommendations. By clicking at
the graph, the connections between these entities
can be explored. The system also shows a short
description for each recommended band/artist.
The system allows two types of controls over
the recommendations: users can adjust the weights
of their items and their friends (initially weighted
by similarity). By changing the weight of your
friend, the system will recognize the change and
calculate your friend overall compatibility match
with you, and in return show whether that friend is
closer to you in term of interest and closeness, or
farther.
4 SOCIAL NETWORKING
SYSTEM FOR COLLECTIVE
INTELLIGENCE
In a collective intelligence environment,
collaborative content can gain more exposure in a
user group that is interested in the subject. Each
user has a connection with another user inside or
outside his group, making it possible for him to
receive suggestion on recommended item from
different perspective, and easy to discuss variety of
topics with people from different views. We can
benefit from this system by connecting users and
encouraging users to contribute in discussion and
share useful materials.
The converse is ‘crowd’ intelligence, typified
by independent persons, often unknown to each
other and always uninfluenced by each other. This
offers a unique approach of information harvesting
for a diverse audience, aim to design a system with
many features, four of which are outlined as
follows (Keller, 2011).
VoiceView application combines social
network with collective intelligence with these
features:
The first is to have effective individual
incentive, organizational structures and
information technology tools.
The second is to pull together distributed
CollectiveIntelligence-HowCollaborativeContentsandSocialMediaChangingtheFaceofDigitalLibrary
351
knowledge within communities that are trying to
solve practical problems combining them into
something useful.
The third is to ensure that error correction
exceeds the rate of error introduction as the system
learns. It means while we are tracking errors that
occurs when the system learns the tasks, we should
gradually improve the system and fix the errors, so
there are some fixes for the remaining errors.
The fourth is to maintain the process
sustainably. It is targeted to handle the information
processing issues which include algorithms and the
associated structures and incentives that make the
algorithm function in the real world.
Among the flowing data and masses
contribution, the system faces challenges to give a
valid value and identify a particular contribution.
First, how we can encourage user contribution. We
can setup a badge system where user will have a
level. And second, how we can filter and track all
the contents and annotations that user sends,
including checking comments. We would apply
points system to encourage users’ contribution.
Each time they refer new friends or upload an
acceptable content, or contribute by sharing useful
comment and annotation, they will receive points.
Once the points reach certain amount, their will get
upgrade to the next level. User with high level will
get more exclusive access to new contents, private
discussion forum, and will get a new badge icon in
their profile.
5 IMPLEMENTATION
In our digital library, multimedia files and text
materials are saved in a web-based database
distributed system, and each material is viewed in
a specific webpage for the material, also known as
item or item webpage. Each of these webpages
contains the picture of an item and explanation,
and, additionally, embbed multimedia content or
references links.
Each item page has annotation suggestion form,
social bookmarking button so users or non-
registered visitors can share links with their own
social media account, and a comment column
where users can discuss about the item and related
topics.
For quality control, it is necessary to authorize
each user input such as comments, files and
annotation suggestions. Admin authorization works
by tracking content and annotation, tracking users’
and groups’ behavior, filtering spam content and
publishing validated inputs. After filtering and
authorizing, admin will publish the item page links
in various social media sites using our own
accounts. Admin creates trailer or samples for
video and audio files, which will be uploaded to
video or audio sharing sites in our own channel.
Table 1: User Collective Intelligence Process.
Users’ role: Contribution: Nature of contribution
A. Users are ‘information
providers’; user supply content and
structure is not specified.
_ Creating blogs
_ Creating wikis
_ Uploading multimedia objects
Time Asynchronize,
unstructured, public
B. Users provide unstructured
responses to existing content
published by the information
provider.
_ Commenting on items (reviews, thoughts, idea)
_ Annotating online texts
_ Uploading multimedia objects
Time Asynchronize,
unstructured, public and
private
C. Users provide structured
responses to existing content
published by the information
provider.
_ Ranking items on a scale
_ Answering multiple-choice opinion surveys
_ Uploading multimedia objects
Time Asynchronize,
structured, private
D. Users create links between
materials
_ Tagging objects
_ Annotating objects with links
_ Sharing links (over social network, email, etc.)
Time Asynchronize,
structured, public
E. Users browsing activities for
personal benefit.
_ Creating personal (private) pages
_ Uploading multimedia objects
_ Tagging objects within a personal space
_ Linking items (private bookmarking)
Time Asynchronize,
unstructured, public
F. Users contribute to our database
-Upload files
Time Asynchronize,
structured, private
WEBIST2013-9thInternationalConferenceonWebInformationSystemsandTechnologies
352
Figure 1: Digital Library System with Collective Intelligence.
Digital library management task list including:
1. Identify user interest and personality
2. Identify users/groups with similar interests
3. Manage discussion
4. Track content trends
5. Suggest user to share links that they like with
their friends
6. Trigger voluntary social campaign
Voluntary social campaign is a recommending
process where user is genuinely interested with
certain subjects, webpage or document and shares
the links in their social media account and by email
to their friends or colleagues. In our digital library,
users can also share links to other users inside their
friends list or inside their group. By attracting
more users, we have more chances to gather more
user-generated content. Users can use discussion
feature, share thoughts in comments section in
each item page. Discussion forum lets users talk
about a particular topic and interact with each
other, express agree or disagree for each other
comment. Users will have a limited space private
message box to send message to their friends. Each
user has a personal page where they can see their
friend list and group list. While users browse
inside digital library and find someone or some
groups with similar interest or intellectual level,
they can join that group, or invite this person in
question to be a friend.
Outside the digital library, users share the links
of their favorite item pages in their social media
accounts. They can also follow our twitter account,
YouTube channel, and promote it to their friends
inside that social media site. If they think their
friends will benefit from our digital library as a
whole, they can share the main page links with
social media accounts or email to their
friends/colleagues. Additionally, we also promote
our own social media accounts by uploading
educative and attractive content in various formats,
such as audio files in iTunes, video in YouTube
channel, introduction to historical items in
SlideShare presentation, pictures of items in Flickr
and Picasa, and publish some free trial documents
as a sample of our content, in some file-sharing
sites.
6 CONCLUSIONS AND FUTURE
WORK
Collective intelligence will provide a resourceful
virtual environment where user can save time and
effort to find necessary information in their field.
Additionally, it is also a platform to discuss and
exchange their thoughts with other scholars and
professionals with experiences in the topic in
question. Our digital library will be a place where
people can find a community to learn and share
knowledge and materials, and their contribution
will benefit every reader.
For future use, we recommend to keep filtering
users that registering. We should also track content,
including content that users upload and comments
that they write inside the item page, and also track
and filter the annotations that they suggest for the
item. Filtering process means to authorize which
content that is topic-related, useful, and written or
mana
g
e
create
submit
Use
r
p
r
o
fil
e
User
Server
Admin
Groups
Friends
Item
Picture
Audios
Video
Article lin
k
Flickr
Facebook
Youtube
Facebook
Twitte
r
AudioArchieve iTunes
Dail
y
Motion
Comments
User submitted
content
CollectiveIntelligence-HowCollaborativeContentsandSocialMediaChangingtheFaceofDigitalLibrary
353
presented properly. Another necessary work is to
update the news, discussion topics and content in
our social media accounts. When there is an
important new content for a particular subject
category in our digital library, we can also put the
news in social media. By updating our social
media news, filtering user-generated content and
encouraging users’ discussion both in our digital
library and in our social media page, we create a
sustainable collective intelligence system.
ACKNOWLEDGEMENTS
This work is supported by National Basic
Research Program of China (973 Program)
No.2011CB302302 and Tsinghua University
Initiative Scientific Research Program.
REFERENCES
Levy, D. M. & Marshall, C. C., 1995. Going digital: A
look at assumptions underlying digital libraries.
Communications of the ACM, 38(4), 77-84. 1995.
Boyd, D. M., & Ellison, N. B., 2007. Social network
sites: Definition, history, and scholarship. Journal of
Computer-Mediated Communication, 2007.
Groh, G., Birnkammerer, S. and Köllhofer, V, 2012.
Social Recommender Systems. Recommender
Systems for the Social Web. Springer (2012). 3–42.
Zann Gill, 2012. User-Driven Collective intelligence
Social Networks as Crowdsourcing Ecosystems. In
Conference on Human Factors in Computing
Systems CHI 2012.
Di Cerbo, F., Dodero, G. and Succi, G., 2008. Social
Networking Technologies for Free-Open Source E-
Learning Systems. In IFIP (International Federation
for Information Processing), 2008.
Aparicio, M, Costa, C, 2012. Collaborative Systems:
Characteristics and Features. In the 30th Annual
International Conference on Design of
Communication (SIGDOC), USA, 2012.
Kwon, G. Y. Lee, Y. S., Kumar, M., 2011. The Tree of
Knowledge: A Localized Collective Intelligence
Tool. In 2011 ACM Conference on Computer
Supported Cooperative Work, CSCW 2011,
Hangzhou, China, March 19-23, 2011.
Ribière, M., Picault, J., Squedin, S., 2010. The sBook:
Towards Social and Personalized Learning
Experiences. In 3rd Workshop on Research advances
in large digital book repositories and complementary
media(Books Online) 2010.
Sarwat, .M, Mokbel, M. F., 2012. RecDB: Towards
DBMS Support for Online Recommender Systems.
In SIGMOD International Conference on
Management of Data, SIGMOD 2012, Scottsdale,
AZ, USA, May 20-24, 2012.
Bostandjiev, S., O’Donovan, J. and Höllerer, T., 2012.
Taste-Weights: A Visual Interactive Hybrid
Recommender System. In the sixth ACM conference
on Recommender Systems. Dublin, Ireland, 2012.
Rachel Keller, 2011. Local Community Intelligence,
Wellbeing and the Potential Role of Web 2.0
Applications. In British Computer Society
Conference on Human-Computer Interaction, BCS-
HCI 2011, Newcastle-upon-Tyne, UK, July 4-8,
2011.
WEBIST2013-9thInternationalConferenceonWebInformationSystemsandTechnologies
354
