The Bio-jETI Framework - User-Level Workflow Design: A Bioinformatics Perspective - page 44

Solving these problems might require the introduction of further computa-
tional steps, for instance a series of conversion services in case of a data type
mismatch. The automatic creation and insertion of such missing service se-
quences is in fact possible with the automatic workflow composition methods
introduced in the next section.
Further on, inter procedural analyses could similarly be integrated via
model checking of context-free/pushdown systems as described in [58, 59, 60].
In fact, it would even become possible to eciently check fork-join parallel
programs with procedures for the popular bitvector problems [153, 278]. This
would allow one to elegantly capture the typical kind of parallelism that ap-
pears in workflows. It is envisaged to integrate these approaches into the
jABC to form a workflow analysis framework in the spirit of the Fixpoint
Analysis Machine [298], which combined a variety of intra- and interproce-
dural methods. Finally, the typical handling of data in workflows often leads
to situations where one and the same object may have many syntactically
different representations. It is therefore planned to investigate the impact of
alias-sensitive analysis methods like the ones based on the Value Flow Graphs
[300].
2.3 Constraint-Driven Workflow Design
The previous section discussed how the continuous evaluation of constraints
via model checking facilitates a constraint- guarded workflow design style.
Constraint- driven workflow design extends the application of constraints
further, systematically using them for high-level descriptions of individual
components and entire workflows, which can then be translated automati-
cally into concrete workflows that conform to the constraints by design. This
section introduces the functionality for (semi-) automatic, constraint-driven
workflow design by which the jABC has been extended in the scope of this
work. More precisely, it describes the PROPHETS (Process Realization and
Optimization Platform using a Human-readable Expression of Temporal-logic
Synthesis) plugin [234], which is the current implementation of the loose pro-
gramming paradigm [178]. It was started by Stefan Naujokat in his Diploma
thesis [232] and has since then been developed further in the scope of different
application projects [233, 180, 179].
In accordance with the loose programming paradigm, PROPHETS sim-
plifies workflow development in order to reach application experts without
programming background. Therefore, the workflow designer neither needs to
model fully executable workflows (as usually necessary) nor to formally spec-
ify a synthesis or planning problem in terms of some first-order or temporal
logic. While behind the scenes the synthesis algorithm naturally requires for-
mal specifications of the synthesis problem, PROPHETS hides this formal
complexity from the user and replaces it by intuitive (graphical) modeling
concepts. Thus, it integrates automatic service composition methodology into