CHI 97 Electronic Publications: Development Consortium
CHI 97 Prev CHI 97 Electronic Publications: Development Consortium Next

An Approach to Evaluation of Software Visualization

Vladimir Averbukh
Inst. for Math. and Mech. of Urals Branch of Russian Academy of Sciences
GSP 384, Inst. Math. Mech. 16. S. Kovalevskoi. Ekaterinburg. Russia. 620219
+7 3432 493254
averbukh@oso.imm.intec.ru

Alexander Konovalov
Inst. for Math. and Mech. of Urals Branch of Russian Academy of Sciences
GSP 384, Inst. Math. Mech. 16. S. Kovalevskoi. Ekaterinburg. Russia. 620219
+7 3432 493248
avkon@oso.imm.intec.ru

Vladislav V. Vorzopov
Urals State University

ABSTRACT

In connection with semiotic aspects of visual languages we define and generalize the content of such conception as visual metaphor, visual language dictionary, visual expressiveness, adequacy in visualization. The experimental system ParaVision should help to search the techniques for evaluating of such characteristics as adequacy in visualization that is as far as a given visual system may satisfy the needs of a given user for solving of a given problem.

Keywords

Visual metaphors, visual expressiveness, adequacy in visualization

© 1997 Copyright on this material is held by the authors.



EXTENDED ABSTRACT

The term "Software Visualization" was suggested in [1] to describe a new domain of Computer Sciences. Software Visualization (SV) includes Visual Programming, Program Visualization and Programming by Demonstration.

In this work we suggest the conception of Visual Languages of SV systems. This conception adds the visual language taxonomy introduced in [2,3]. Under this category, we consider visual programming languages, languages of program visualization systems, visual and iconic languages used in programming by demonstration, visual database querying languages, and languages of iconic interfaces.

We define and generalize the content of these visual languages according to various semiotic aspects: visual metaphor, visual language dictionary, visual expressiveness, and adequacy of visualization.

In the general case the significant unit of the visual language can be the following:

The significant unit of the visual language may be of the complex nature. The important language components are the changes of its characteristics according to the language grammar.

In this paper we consider the visual metaphor as the system of similarities putting a some set of images to a correspondence with concepts and objects of a given application domain.

In our previous work [4,5] we have used the term "visual expressiveness" for describing of some aspects of the visual language pragmatics. To evaluate the pragmatics characteristics we considered such parameters as perception speed, perception uniqueness, fatigue level, and the availability (or lack) of aesthetic and emotional satisfaction. To check these ideas we created two simple visual system prototypes where each had analogous semantics but based on different visual metaphors. Also we had offered the methods of the formalized description of the characteristics of visual languages and visual metaphors.

Our last area of research is linked with problems of visualization for parallel computation, a domain in which visualization can be most useful and effective. Just parallel computations need effective visual metaphors for visualization systems.

The experimental system ParaVision should aid in searching techniques for evaluating such characteristics as adequacy in visualization -- that is, the degree to which a given visual system satisfies the needs of a given user for solving a given problem. The model problem is a popular problem of parallel programming - The Dining Philosophers problem [6].

There are three visual metaphors offered to users: natural, symbolic, and abstract. In the natural metaphor, philosophers, forks, etc., are represented by natural images. In the symbolic metaphor, philosophers are represented by circles (similar to the representations in [7]). In the abstract metaphor, idle, active and waiting-for-lock times are represented by Gantt charts. We then analyze statistics gathered for each user in order to evaluate adequacy in visualization. The techniques for these experiments are currently under development. We suppose that ParaVision users will be Software Engineering students and we hope to begin experimental use of ParaVision during this year.

REFERENCES

1. Price B.A., Small I.S., Baecker R.M. A Taxonomy of Software Visualization. In: IEEE Computer Society Press Reprint. 1992.

2. Shu N.C. Visual Programming Languages: A Perspective and Dimension Analysis. In: Visual Languages. New York. Plenum Publishing Corporation. 1986. p. 11-34.

3. Chang S.-K. Visual Languages: A Tutorial and Survey. Visualization in Programming. (Lecture Notes in Computer Science 282). Berlin. Springer-Verlag. 1987. p. 1-23.

4. . Averbukh V.L., Tarskikh I.V. Language Expressiveness in Software Visualization Systems. In: "East-West". International Conference on Human-Computer Interaction EWHCI'93. Moscow. Russia, 3-7 August, 1993". Proceedings. ICSTI. Moscow. 1993. V.II, pp. 5-8.

5. Averbukh V.L., Konovalov A.V., Tarskikh I.V., Vorzopov V.V. Analyses of Visual Metaphors and Languages. Toward Prototyping of Software Visualizing Systems. In: "East-West". International Conference on Human-Computer Interaction EWHCI'94. St.-Peterburg, Russia, 2-6 August, 1994". Proceedings. ICSTI. Moscow. 1994. V.I, p. 244-254.

6. Hoare C.A.R. Communicating Sequential Processes. Printice-Hall. Englewood Cliffs. 1985.

7. Price B.A., Baecker R.M. The Automatic Animation of Concurrent Programs. In: 1-st Moscow International Workshop Proceedings HCI'91 Moscow 5-9 August 1991. ICSTI. Moscow. 1991 pp. 128-137.


CHI 97 Prev CHI 97 Electronic Publications: Development Consortium Next

CHI 97 Electronic Publications: Development Consortium