2020/09/01 by Philipp Meschenmoser, Juri Buchmüller, Meschenmoser, Philipp +7 · 1 citation
Environmental Science · Agricultural and Biological Sciences · #Species Distribution and Climate Change #Bat Biology and Ecology Studies #Amphibian and Reptile Biology
paper · pdf · doi:10.48550/arxiv.2009.00548
Segmenting biologging time series of animals on multiple temporal scales is\nan essential step that requires complex techniques with careful\nparameterization and possibly cross-domain expertise. Yet, there is a lack of\nvisual-interactive tools that strongly support such multi-scale segmentation.\nTo close this gap, we present our MultiSegVA platform for interactively\ndefining segmentation techniques and parameters on multiple temporal scales.\nMultiSegVA primarily contributes tailored, visual-interactive means and visual\nanalytics paradigms for segmenting unlabeled time series on multiple scales.\nFurther, to flexibly compose the multi-scale segmentation, the platform\ncontributes a new visual query language that links a variety of segmentation\ntechniques. To illustrate our approach, we present a domain-oriented set of\nsegmentation techniques derived in collaboration with movement ecologists. We\ndemonstrate the applicability and usefulness of MultiSegVA in two real-world\nuse cases from movement ecology, related to behavior analysis after\nenvironment-aware segmentation, and after progressive clustering. Expert\nfeedback from movement ecologists shows the effectiveness of tailored\nvisual-interactive means and visual analytics paradigms at segmenting\nmulti-scale data, enabling them to perform semantically meaningful analyses. A\nthird use case demonstrates that MultiSegVA is generalizable to other domains.\n