Information Technology Reference
In-Depth Information
ServiceMap
Map and GPS systems
Q1
: what other services do people
who use these services also use?
Q2
: what is an operation chain
between two services/operations?
Q1
': what other places do people
who visit these places also visit?
Q2
': what is a route between two
places?
Figure 8.11
The analogy between ServiceMap and Map/GPS systems.
8.3.2 ServiceMap Approach
As mentioned earlier, Q1 and Q2 are not isolated but raised in different
stages in an
in-silico
experiment. Q1 is usually raised when an experi-
ment is at its conceptual level, and scientists need to figure out the
available resources to leverage. Q2 is raised in a later stage when the
routine of an experiment needs to be concretized at the operational
level. It is also a common practice to iterate over these two stages due to
the explorative nature of bioinformatics experiments. Inspired by this
requirement from multiple user communities including caBIG, we
present ServiceMap as a framework to address both Q1 and Q2, in a
holistic manner.
As illustrated in Figure 8.12 , ServiceMap leverages the two disjoint
networks, that is, an undirected
workflow-service network
and a
directed
operation network
, both derived in the network analysis as
shown in Section 8.2.
In the undirected workflow-service network, nodes are either work-
flows or services, and edges represent the inclusive relations between
them. In the directed operation network, nodes are operations in services,
and a directed edge represents a data link between two operations in
some workflow. While Section 8.2 focuses on how to build and calculate
the metrics of these networks, ServiceMap focuses on algorithms on
these networks to answer Q1 and Q2. A brief summary of these
algorithms is as follows. To answer Q1, we derive the frequent item
sets and association rules in the workflow-service network, and recom-
mend relevant services in a given context, that is, existing services in a
