Information Technology Reference
In-Depth Information
Fig. 5.17
Correlation graph obtained from Microsoft query logs, considering all relevant search
terms
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Cluster 5
complexity
evolution
evolutionary
chaos
cognition
biology
theory
behavior
markets
genetics
agent
systems
visualization
organization
information
community
networks
ai
emergence
neural
algorithms
ecology
math
computing
optimization
computation
environment
mathematics
physics
economics
art
science
simulation
dynamics
nonlinear
computational
ecosystem
Fig. 5.18 Optimal partition into clusters, obtained from the Microsoft query data, when the top
200 edges are considered, smallest cluster not shown. The resulting partition has a Q
=
0
.
536.
However, nine terms were assigned to their own cluster, thus basically left unclassified
5.6.2
Discussion of the Results from the Query Log Data
and Comparison
When comparing the graphs in Figs. 5.11 and 5.16 (i.e. the ones which only
depict the relations to the central term 'complexity') an important difference can be
observed. While the graph in Fig. 5.11 , based on collaborative tagging data, shows
48 terms related to complexity, the one in Fig. 5.16 , based on query log data, shows
 
Search WWH ::




Custom Search