BraidFlow: A Flow-annotated Dataset of Kumihimo Braidmaking Activity

Abstract

Entering a cognitive state of fow is a natural response of the mind that allows people to fully concentrate and cope with tedious, and often repetitive tasks. Understanding how to trigger or sustain fow remains limited by retrospective surveys, presenting a need to better document fow. Through a validation study, we frst establish braidmaking as a fow-inducing task. We then study how braidmaking can be used to unpack the experience of fow on a moment-by-moment basis. Using an instrumented Kumihimo braidmaking tool and of-the-shelf biosignal wristbands, we record the experiences of 24 users engaged in 3 diferent braidmaking tasks. Feature vectors motivated from fow literature were extracted from activity data (IMU, EMG, EDA, heart rate, skin temperature, braiding telemetry) and annotated with Flow Short Scale (FSS) scores. Together, this dataset and data-capture system form the frst openaccess and holistic platform for mining fow data and synthesizing fow-aware design principles.