Koldijk, SaskiaBernard, JürgenRuppert, TobiasKohlhammer, JörnNeerincx, MarkKraaij, WesselE. Bertini and J. Kennedy and E. Puppo2015-05-242015-05-242015https://doi.org/10.2312/eurovisshort.20151129Stress in working environments is a recent concern. We see potential in collecting sensor data to detect patterns in work behavior with potential danger to well-being. In this paper, we describe how we applied visual analytics to a work behavior dataset, containing information on facial expressions, postures, computer interactions, physiology and subjective experience. The challenge is to interpret this multi-modal low level sensor data. In this work, we alternate between automatic analysis procedures and data visualization. Our aim is twofold: 1) to research the relations of various sensor features with (stress related) mental states, and 2) to develop suitable visualization methods for insight into a large amount of behavioral data. Our most important insight is that people differ a lot in their (stress related) work behavior, which has to be taken into account in the analyses and visualizations.I.5.4 [Pattern recognition]ApplicationsSignal processingH.5.0 [Information Interfaces and Presentation]General10.2312/eurovisshort.2015112979-83