Analysis of Evoked Emotions in Extremist YouTube Videos Through Kansei Evaluation

TLDR: In this article, the authors report on factor analysis of 62 emotional descriptors rated by 80 university students after having watched 20 extremist YouTube videos during a Kansei evaluation and successfully ascertained significant emotion descriptors are successfully determined wherein 27 descriptors were retained in 3 factors; offensive, intrigue and awkward.

Abstract: Extremist YouTube videos are always associated with the ‘Dark side’ and potentially bring negative influence on younger viewers. These videos are deemed inappropriate for containing violent, has since triggered interest of academia and other stakeholders to study on its content. Extensive body of research mostly emphasizes on extremists’ messages, modus operandi, production features and may overlook on emotion perspective of the viewers. Viewing the videos can evoke specific emotional response in the viewers. Emotional response can be described in specific sets of adjective words or sentences which can be referred as emotional descriptors. This paper will report on factor analysis of 62 emotional descriptors rated by 80 university students after having watched 20 extremist YouTube videos during a Kansei evaluation. Significant emotional descriptors are successfully ascertained wherein 27 descriptors are retained in 3 factors; offensive, intrigue and awkward. The remaining 35 descriptors can be ignored as the proportions of variability explained by the remaining factors are close to zero and can be considered insignificant. The finding for ‘offensive’ and ‘awkward’ is expected since majority of the descriptors are of negative affect in nature. Rather interesting is the finding about ‘intrigue’ as a factor since it contains positive affect in circumstance which commonly perceived as negative. That however, adds to novelty for this research. Together the result shall benefit academia and other stakeholders in understanding evoked emotions in extremist YouTube videos and future work on affective video classification.