Mining Bodily Cues to Deception

Abstract

A significant body of research has investigated potential correlates of deception and bodily behavior. The vast majority of these studies consider discrete, subjectively coded bodily movements such as specific hand or head gestures. Such studies fail to consider quantitative aspects of body movement such as the precise movement direction, magnitude and timing. In this paper, we employ an innovative data mining approach to systematically study bodily correlates of deception. We re-analyze motion capture data from a previously published deception study, and experiment with different data coding options. We report how deception detection rates are affected by variables such as body part, the coding of the pose and movement, the length of the observation, and the amount of measurement noise. Our results demonstrate the feasibility of a data mining approach, with detection rates above 65%, significantly outperforming human judgement (52.80%). Owing to the systematic analysis, our analyses allow for an understanding of the importance of various coding factor. Moreover, we can reconcile seemingly discrepant findings in previous research. Our approach highlights the merits of data-driven research to support the validation and development of deception theory.

Keywords: Body motion; Data mining; Deception; Motion capture; Movement analysis.

Fig. 1

From left to right: a Location of the 23 joints. Root joint in red. Body parts are indicated with different colors. b–e Schematic visualization of the four feature types: movement, joint angle, joint distance, and symmetry

Fig. 2

Classification scores in percentages for different window sizes (in seconds), obtained using all, stat-95 and stat-99 features

Fig. 3

Classification scores in percentages for added noise with different factors r, obtained using all, stat-95, and stat-99 features. Scores are averaged over 100 repetitions

Fig. 4

Classification scores in percentages for different numbers of features in decreasing order of statistical difference, obtained with a window size of 2.5 min

Fig. 5

Visual representation of the percentage of features selected in stat-95 and stat-99. Darker colors correspond to higher percentages