Rapid Visualization Development based on Visual Programming

Information

• Publication Type: Master Thesis
• Workgroup(s)/Project(s):
  • Visualization Group
• Date: 2011
• First Supervisor: Martin Haidacher
• Keywords: Rapid Development, Dataflow Programming, Visual Programming, Visualization

Abstract

Over the years, many visualization tools and applications were developed for specific fields of science or business. Staying in the alcove of their field, these are highly suited and optimized for visualizing specific data, with the drawback of not being flexible enough to extend or alter these visualizations for other purposes.

Often, customers of such visualization packages cannot extend them, to fit their needs, especially if the software is closed source. But even using open source software does not solve this problem efficiently, since to extend the software, costumers need to have programming skills and are often forced to reimplement algorithms or visualizations which already exist, making rapid development impossible.

The goal of this thesis is to develop a dataflow visual programming language (DFVPL) and a visual editor for the rapid development of visualizations. With this programming language, called OpenInsightExplorer, users can develop visualizations by connecting graphical representations of modules rather than writing source code. Each module represents a part of a visualization program. Modules are designed to function as an independent black box and they start to operate as soon as data is sent to them. This black box design and execution model allows to reuse modules more frequently and simplifies their development.

This programming language and its editor run platform independently to reach a high number of potential programmers, respectively users, to develop visualizations, since they are not bound to a specific platform. It is extendable, by means of self developed modules and data types to extend the language. Programming and editing a visualization is easy and fast, even for people with only little programming experience. The production cycle of the development of visualizations is reduced to a minimum. This is achieved by reusing and combining existing modules.

The usability of the programming language was evaluated by implementing two example visualizations with it. Each example originates from different areas of visualization, therefore demanding different data types, data transformation tasks and rendering.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@mastersthesis{stehno-2011-RVD,
  title =      "Rapid Visualization Development based on Visual Programming",
  author =     "Benedikt Stehno",
  year =       "2011",
  abstract =   "Over the years, many visualization tools and applications
               were developed for specific fields of science or business.
               Staying in the alcove of their field, these are highly
               suited and optimized for visualizing specific data, with the
               drawback of not being flexible enough to extend or alter
               these visualizations for other purposes.   Often, customers
               of such visualization packages cannot extend them, to fit
               their needs, especially if the software is closed source.
               But even using open source software does not solve this
               problem efficiently, since to extend the software, costumers
               need to have programming skills and are often forced to
               reimplement algorithms or visualizations which already
               exist, making rapid development impossible.   The goal of
               this thesis is to develop a dataflow visual programming
               language (DFVPL) and a visual editor for the rapid
               development of visualizations. With this programming
               language, called OpenInsightExplorer, users can develop
               visualizations by connecting graphical representations of
               modules rather than writing source code. Each module
               represents a part of a visualization program. Modules are
               designed to function as an independent black box and they
               start to operate as soon as data is sent to them. This black
               box design and execution model allows to reuse modules more
               frequently and simplifies their development.   This
               programming language and its editor run platform
               independently to reach a high number of potential
               programmers, respectively users, to develop visualizations,
               since they are not bound to a specific platform. It is
               extendable, by means of self developed modules and data
               types to extend the language. Programming and editing a
               visualization is easy and fast, even for people with only
               little programming experience. The production cycle of the
               development of visualizations is reduced to a minimum. This
               is achieved by reusing and combining existing modules.   The
               usability of the programming language was evaluated by
               implementing two example visualizations with it. Each
               example originates from different areas of visualization,
               therefore demanding different data types, data
               transformation tasks and rendering.",
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Institute of Computer Graphics and Algorithms, Vienna
               University of Technology ",
  keywords =   "Rapid Development, Dataflow Programming, Visual Programming,
               Visualization",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2011/stehno-2011-RVD/",
}