At the 17. IEEE Pacific Visualization Conference (PacificVis 2024), Franziska Becker and Tanja Blascheck received the VisNotes Honorable Mention Award for their short paper “TimeSeriesMaker: Interactive Time series Composition in No Time”. Congratulations to them!
PacificVis 2024 took place from April 23 to 26 in Tokyo, Japan. The aim of PacificVis is to promote visualization research and technologies, especially in the Asian-Pacific region. The Visualization Notes (VisNotes) track is a short paper track of the conference that aims to encourage young researchers in particular to present their work and discuss it with the participants.
Publikation
F. Becker and T. Blascheck, “TimeSeriesMaker: Interactive Time Series Composition in No Time,” in 2024 IEEE 17th Pacific Visualization Symposium (PacificVis), in 2024 IEEE 17th Pacific Visualization Symposium (PacificVis). IEEE Computer Society, 2024.
Abstract
TimeSeriesMaker is an open-source application to visually compose time series data in an intuitive and shareable manner. Visualization researchers often use time series data in studies about perceptual or cognitive phenomena and many other contexts. However, finding or generating time series data that fits a given scenario is not always easy. Using a component-based architecture, TimeSeriesMaker allows analysts to compose time series data with complex patterns by combining different components, such as noise, a linear trend or a seasonal pattern. An interactive compositor tree of these components lets analysts explore their combinations using different operators. We support reproducibility and transparency by including functionalities that allow analysts to export and share their configuration, which others can use to reload and modify the same time series. In a qualitative online study with visualization researchers, we found that our approach enables them to create a time series based on an example image or their own requirements. However, system usability could be further improved when interacting with the compositor tree. TimeSeriesMaker can be found here: https://unistuttgart-visus.github.io/time-series-maker/.