Automated Composition of Scientific Workflows: A Case Study on Geographic Data Manipulation

V. Kasalica, A.L. Lamprecht

    Research output: Contribution to conferencePosterAcademic

    Abstract

    We created a semantic domain model for a geographic data manipulation scenario (the creation of a topographic map of the Netherlands) in order to demonstrate what automated workflow composition can accomplish once suitable technical vocabularies and tool annotations are available. The classification of operations and types in the ontology is essential for effective usage of the automated composition framework, as it provides abstract terminology that the user can use to describe workflow at an abstract level. Most of the annotated tools in this case study belong to the GMT (Generic Mapping Tools). We then used the PROPHETS loose programming framework in order to specify and synthesize an executable workflow. The user provides the framework with constraints that describe desirable or undesirable features of the intended workflow (e.g. "use tool to draw water", "don’t use 3D plotting"). The framework translates this specification into a set of temporal logic constraints that are used as an input for the synthesis algorithm, together with the ontologies and tool annotations. The result of the synthesis is a workflow (or a set of different possible workflows) that fulfills the specification, allocating an annotated tool or a sequence of tools for each of the abstractions (e.g. "draw water" will be substituted with an actual tool).
    Original languageEnglish
    DOIs
    Publication statusPublished - 2018
    EventIEEE eScience 2018 - Amsterdam, Netherlands
    Duration: 29 Oct 20181 Nov 2018
    https://www.escience2018.com/

    Conference

    ConferenceIEEE eScience 2018
    Country/TerritoryNetherlands
    CityAmsterdam
    Period29/10/181/11/18
    Internet address

    Fingerprint

    Dive into the research topics of 'Automated Composition of Scientific Workflows: A Case Study on Geographic Data Manipulation'. Together they form a unique fingerprint.

    Cite this