Preface to IEEE VAST 2020 Conference Track and VAST Challenge

Abstract

This is the 15th edition of IEEE Visual Analytics Science and Technology (VAST). Begun in 2006 as an IEEE Symposium at VIS, it is now in its 11th year as an IEEE Conference. It continues to be the leading forum for Visual Analytics research, defined as the science of analytical reasoning supported by interactive visual interfaces. VAST represents research pushing the boundaries of the state of the art in theory and foundations of visual data analysis, techniques and algorithms, empirical and design studies, as well as systems and applications. VAST in 2020 continues to feature its successful conference paper track, in addition to the TVCG paper track. The goal of this track is to increase the diversity of Visual Analytics applications and to better support participation of interdisciplinary researchers. It provides innovative advances and applications in Visual Analytics. The VAST 2020 Program Committee comprised 59 senior experts from the field. 210 complete submissions entered the two-stage review cycle, from which VAST eventually accepted 51 papers for the TVCG track, and 10 for the conference track. The conference track papers are published as part of the VIS USB proceedings, and submitted to the IEEE Digital Library for archival publishing. The accepted papers contribute interesting, timely ideas and results to the VAST 2020 conference sessions on Fairness and AI, Interactive Machine Learning, Text Analysis, Graphs, Evaluation and Theory, as well as Applications. Now in its 15th year, the IEEE VAST Challenge continues to pose new challenges to the visual analytics research community to encourage innovation in interactive visual representation, data transformation, and analytical reasoning. This year’s three minichallenges centered around a global internet outage, and tested participants’ abilities to explore and compare graphs, draw conclusions from poorly classified images, and to design a future visual analytic environment. The datasets and submissions are archived in the Visual Analytics Benchmark Repository (https://www.cs.umd.edu/hcil/varepository/), and papers for several submissions are published as part of the VIS USB proceedings. This year’s submissions illustrate the power of combining machine learning and interactive visualization to gain insight into complex problems.