Analysis of emotional descriptors for video-watching experience through Kansei evaluation

• Published in: Advanced Science Letters
• Main Author: Rosli R.M.; Lokman A.M.
• Format: Article
• Language: English
• Published: American Scientific Publishers 2017
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85023741947&doi=10.1166%2fasl.2017.8321&partnerID=40&md5=4862d68737263d3027b49f583a5c6754

Watching videos bring about experience that touches the heart and minds of viewers. It is literally an emotional experience involving psychological as well as physical reaction. Verbally describing an emotional experience involves adjective words or sentences that represent the emotional state and its intensity. For example, watching a baby laughing unstoppable when being tickled causes a viewer to have positive emotional experience by feeling so happy. While the other viewer may experience same feeling but at different intensity or even totally opposite feeling. The specific sets of adjective words or sentences in describing emotional experience or the emotion can be referred as emotional descriptors. There are many previous research highlighted on different emotional descriptors and its usage to measure affective response in various domains. During emotional descriptors selection for a particular study, researcher will have to weigh some considerations and select those that are most relevant to their research. However, most video content analysis research are adopting generic and basic descriptors. This paper will report on analysis to reduce 60 emotional descriptors adopted from PANAS-X scale being rated by 35 students upon watching 5 video during a Kansei evaluation. Reduction of the descriptors based on correlation values are successfully achieved when it resulted in 32 descriptors inclusive of 4 new cluster descriptors that best describe about video-watching experience. It is also found that the new cluster descriptors correctly described four of the video specimens. The finding concurs with the need to select descriptors that best fit in describing a particular research domain for better result. It will benefit academia and other stakeholders towards affective video content analysis in general and accessing affective response pertaining to video-watching experience specifically. © 2017 American Scientific Publishers All rights reserved.