Jamshid Rahimi. Local clustering effects in online response networks. Master's Degree(Information Technology). King Mongkut's University of Technology North Bangkok. Central Library. : King Mongkut's University of Technology North Bangkok, 2009.
Local clustering effects in online response networks
Abstract:
Online Learning Networks have become an important new learning form and
yet relatively little is known about the underlying mechanisms. In an attempt to
address this knowledge gap, a particular subset of online networks online response
networks - is investigated. We introduce hypotheses related to the underlying
mechanism of local clustering in online learning networks. The hypotheses are based
on social network theory and tested by comparing with social networks and Random
Graph Models and are expressed in terms of the abundance of certain structures
(motifs) in response graphs representing the online learning networks.
In this work we report two new results. The results are based on analyzing a
sample of 94 observed online learning networks as well as a standard set of social
networks. The first result is that in typical online learning networks, and unlike social
networks, local clustering can be explained by actors choosing their response partners
at random, subject to their inherent total response capacities and reciprocity; no extra
mechanism is required. This result was obtained by comparing the local clustering in
the observed networks with ensemble networks generated by Markov Chain Monte
Carlo (MCMC) simulation of constrained random graph models. The second result is
that in online learning local clustering is correlated with reciprocity, the more
reciprocity, the more they are clustered. But in social networks no significant
correlation was seen. Local clustering is a result of an external mechanism.
Several mechanisms are suggested in the literature as the source for the
phenomenon of clustering in social networks. The absence of these mechanisms in
online learning networks, and implications of the results to the design of online
learning networks are briefly discussed.