An Evaluation of End-User Interfaces of Scientific Workflow Management Systems

No Thumbnail Available

Date

2005-01-07

Journal Title

Series/Report No.

Journal ISSN

Volume Title

Publisher

Abstract

Scientific research is often exploratory in nature. As technology advances, more steps of such work may become automated. Data used and produced in such experiments becomes increasingly complex and heterogeneous. In such scenarios, a need arises for a tool (or a set of tools) that allows dynamic modeling and (semi) automatic construction and integration of problem solving flows (scientific workflows) using various, often network-based, components. Various Workflow Management Systems (WFMSs) have been developed that allow modeling, construction, and execution of business-oriented workflows. Business workflows are, however, different from scientific WFMSs. They often are less dynamic and evolving in nature. Scientific workflows tend to change more frequently and may involve very voluminous data-translations. In addition, while business workflows tend to be constructed by professional software and business flow engineers, scientific workflows are often constructed by scientists themselves — experts in their domains, but not necessarily experts in information technology, the software or networking domains in which the tools and workflows operate. Therefore, the two cases may require considerably different interfaces and end-user robustness both during the construction stage of the workflows and during their execution. Currently, a number of issues related to support of scientific workflows remain open. One is the issue of their graphical and other user interfaces. SciDAC?s Scientific Data Management (SDM) initiative is developing a framework that allows scientists to manage data in a more efficient manner, using tools that help them create and manage scientific workflows that use network-based (web) services. This work is part of that effort and it focuses on a comparative analysis of graphical user interfaces of the SDM framework with some other available solutions. An approach for evaluating SWFMS end-user interfacing is presented. More than 20 criteria, based on HCI literature and the findings from the prototype SWMS developed as part of SDM research, are presented. The criteria are used to compare and discuss interfaces of the current version of the SDM SWFMS — the Scientific Process Automation system based on Ptolemy II framework, with some other systems such as SCIRun, Enhydra JaWE, and Taverna. The study finds that although a lot of effort has been put into creating user-friendly workflow systems, the complexity of the visual component is still overwhelming and unintuitive at times — even to computer science students. Many systems that we discuss here employ interesting metaphors, but as the complexity of the underlying layers increases, most visual components also become more complex and more difficult to use. Special care should be taken in presenting new or enhanced functionality to the user for the system to be well accepted. We also observe that most scientific workflow systems that may be used in the life sciences domain do not support satisfactorily decision-based construction and execution of workflows. At this time, most of these systems have difficulty in interoperating with other supporting systems. Verification, Validation and Fault-Tolerance also appear to be a challenge.

Description

Keywords

Evaluation, Graphical User Interfaces, Scientific Workflows, Workflow Systems

Citation

Degree

MS

Discipline

Computer Science

Collections