Database Reference
In-Depth Information
12.6.5 Enabling Multidisciplinary and Multiscale
Applications Using Provenance
As we have seen, provenance plays an important role in enabling scientific
applications. In particular, those applications that use a variety of heteroge-
neous data sources or computational resources benefit. Scientific problems are
increasingly becoming multidisciplinary and multiscale. For example, biomedi-
cal applications may combine the results of chemistry simulations of molecular
interactions with data about tissues and other organs. 90
To allow the provenance of data to be determined across boundaries of scale,
discipline, and technologies, there is a need for an interoperability layer be-
tween systems. One proposed interoperability specification is the Open Prove-
nance Model. 55 This model provides a good outline of what the community
developing provenance technologies believes are the core constituents of a
provenance graph. We thus use a graphical representation drawn from the
Open Provenance Model to illustrate a concrete provenance graph, as shown
in Figure 12.5.
In the representation we adopt, nodes of the graph consist of two entities: ar-
tifacts and processes. In the context of scientific workflows as considered here,
artifacts are immutable pieces of data, whereas processes are transformations
that produce and consume artifacts. Artifacts are represented as circles, and
processes are denoted by boxes.
Param
Atlas
×
Graphic
Atlas
×
Slice
Atlas
Image
Softmean
Slicer
Convert
Atlas
Header
Legend:
WasGeneratedBy(R)
P
A
Used(R)
A
P
Figure 12.5
Provenance graph for the Provenance Challenge Workflow.
Search WWH ::




Custom Search