what-when-how
In Depth Tutorials and Information
16.5 MacroscopicStudyofSocialNetworks[3]
Since the 1990s many researchers have been studying social network theory due to its
rapid growth. In the earlier phase, researchers focused on small networks in various
computer-supported collaborative learning (CSCL) environments. heir network
investigations were exhaustive and based on assumptions of seeing the social network
as a unirelational network. Unfortunately, their study results are incomplete since
their analyses were based on short-range networks. Several important characteristics
of social networks (Internet, phone call networks, World Wide Web, country road
maps, electricity transmission network, etc.) are left out of such networks, such as
detection of the network type, formation of giant clusters, and network properties.
It is unquestionable that the CSCL environments offer many applications to the
user. Web-based/online discussion forum, e-mail, chat room, NetMeeting, video
calls, etc., are some of the tools that are accessible to any member of the network.
However, the actual study focuses on Web-based discussion forums because all par-
ticipants' communication information can be recorded and messages can be recov-
ered easily without “additional efforts for hardware or software modification.”
Suppose that
N
is a set of nodes and
R
the set of relations between the nodes
or objects. From this assumption, we can come up with another set called
G(N,R)
,
which is basically a mathematical model of a network. If
P
represents a set of partic-
ipants and
M
a set of messages posted in a discussion forum, we can build another
network called
N
= {
P,M
}
,.
which is the network for the discussion forum. We
assume that the discussion is engaged by one participant (a single node), and this is
what usually occurs in real life. Hence, let us call
S
the relationship set that means
“who submits/posts that message.” Moreover, the member who starts the conversa-
tion may post many messages or topics, but we do not assume that many partici-
pants post starting discussion messages since forum discussions do not work that
way. According to the article [3], it is a one-to-many mapping from
P
to
M
. Let us
consider the basic network
G
b
({
P,M
},.S)
, where there is no intrarelation between
elements inside the set
P
or the set
M
. hen the question is, what is the right way
to allow communication within elements in set
P
and what are the implications of
studying those linkages in order to understand CSCL?
For a better understanding of the questions raised above, we used an idea of
Newman, a researcher who works on several large-scale networks (up to 1.5 mil-
lion nodes). In fact, he stated that two researchers are socially linked if they have
published at least one paper together. In our online discussion forum case, there
is a always a member who starts the conversation or posts an issue, and the other
participants give their points of view. From that point, it can be said that the topic
starter and the repliers are socially linked since they have been reciprocally exchang-
ing ideas. hen, we can introduce another type of network called the collaborative
learning network
G
c
(
P,M
t
),
where
M
t
is the relationship set representing all the
submissions of messages in order to show the existing co-discussion of a specific
topic by various members.
