To improve the future, we must reflect on our past.
The IEEE VIS Test of Time Award is an accolade given to recognize articles published at previous conferences whose contents are still vibrant and useful today and have had a major impact and influence within and beyond the visualization community.
By making the awards at the conference opening we hope to encourage researchers to aim to produce work that is forward looking and has transformational potential. We’re trying to build on our heritage to establish an ambitious future by making it clear at the conference opening that we want participants to aspire to be writing the papers that will be relevant in 10 and 20 years.
Papers are selected for each of the three conferences (VAST, InfoVis and SciVis) by Test of Time Awards panels. These panels are appointed by each conference Steering Committee.
The decisions are based on objective measures such as the numbers of citations, and more subjective ones such as the quality and longevity and influence of ideas, outreach, uptake and effect not only in the research community, but also within application domains and visualization practice.
VAST currently considers conferences 10 years prior to the current conference.
InfoVis currently considers conferences 10 and 20 years prior to the current conference.
SciVis currently considers 15 and 25 years prior to the current conference.
VAST 2009: 10 Years Test of Time Award
Parallel tag clouds to explore and analyze faceted text corpora
Authors: Christopher Collins, Fernanda B. Viegas, and Martin Wattenberg
https://doi.org/10.1109/VAST.2009.5333443
In this 2009 VAST paper, co-authors Christopher Collins, Fernanda B. Viegas, and Martin Wattenberg investigated such a problem and presented a new visualization technique called Parallel Tag Clouds, successfully visualizing both terms and facets relationships.
This elegant solution has influenced many subsequent research activities and applications dealing with this problem, as demonstrated by the overall number and the increasing trend of citations in the last ten years and by the different research and application fields that today still refer to this 2009 proposal, like applications in the domain of Financial Engineering, Data Management, or Multimedia Management and Retrieval.
Committee: Giuseppe Santucci (chair), Remco Chang, Brian Fisher
InfoVis
1999: Cluster and calendar based visualization of time series data
Authors: J.J. Van Wijk, E.R. Van Selow
https://doi.org/10.1109/INFVIS.1999.801851
Cluster and calendar based visualization of time series data combines time-series cluster analysis with a novel calendar-based visualization scheme to aid the analysis of univariate time-series data. The work’s contributions lie both in its principled combination of data mining and visualization methods, motivated by a real-world domain problem — thus serving as an example for what would become the discipline of visual analytics — and in its contribution of a straightforward and effective calendar visualization scheme that has been adopted by subsequent visualization tools.
The premise of breaking down time-series data into clusters and visualizing those is important and influential, and represents one of the earliest combinations of cluster analysis and visualization. The InfoVis and VAST communities continue to make use of this approach as it can be effective for turning an overwhelming amount of data into digestible chunks.
The paper also includes an interesting and formative redesign process from a 3D representation to a more informative 2D one. The energy conservation application provides a compelling motivation and real-world grounding for the technique contributions. Overall, this work serves as an early exemplar of research that integrates theory/formalism and design/practice.
Committee: Jeff Heer, Fernanda Viegas, Niklas Elmqvist
2009: A Nested Model for Visualization Design and Validation
Authors: Tamara Munzner
https://doi.org/10.1109/TVCG.2009.111
The “Nested Model for Visualization Design and Validation” paper is useful for anyone planning, conducting or assessing visualization research. At its core, it asks a triple of very simple but important methodological questions that value consistency and coherence:
- what assertion is your research paper making?
- what would it take for that assertion to be credible?
- does your paper show this to be the case?
It tackles these big issues comprehensively and pragmatically, showing us how to do applied visualization research appropriately. In doing so, the paper has changed the way in which we think about validity in visualization research, and has been influential beyond our domain.
As a theoretical research paper “Nested Model” is very clear in objectives, claims, description of the model itself, and use of a series of complementary examples as test cases. It reflects very well on its scope, assumptions, limitations, and implications, guiding visualization researchers in the wide range of papers that we write.
The paper has the stated “hope of creating a more unified visualization literature where [problem driven and technique driven research] interleave usefully”. It has enabled the visualization community to make significant progress in achieving this by developing an approach that is the basis for much of the research that we conduct today. Maybe, after a decade of influence, “A Nested Model” is “The Nested Model”.
Committee: Jason Dykes, Danyel Fisher, Jean-Daniel Fekete
SciVis
1994: An evaluation of reconstruction filters for volume rendering
Authors: Stephen Marschner and Richard Lobb
https://doi.org/10.1109/VISUAL.1994.346331
Interpolation underpins all volume rendering algorithms and the choice of an interpolation scheme can have a dramatic effect on image quality. This paper was one of the first to analyze the process of interpolation and present it as a formal basis in 3D signal processing theory. The paper also presented a set of metrics to help quantify the characteristics various filters in use and artifacts resulting from those filters.
Committee: Min Chen, Alex Pang, Deborah Silver
2004: Simplifying Flexible Isosurfaces Using local Geometric Measures
Authors: Hamish Carr, Jack Snoeyink, Michiel van de Panne
https://doi.org/10.1109/VISUAL.2004.96
The contour tree is a graph based representation that displays how the topology of a level set changes with the scalar value. It is a very useful depiction that describes the structure of the scalar field. The paper, Simplifying Felxible Isosurfaces Using Geometric Measures, helped popularize the concept of the contour tree and showed how it could be used for guided interactive data exploration especially for visualizing flexible isosurfaces.
Committee: Min Chen, Alex Pang, Deborah Silver