A paper by Professor SHIRAHAMA Naruki of Faculty of Data Science has been published in the Journal of Advanced Computational Intelligence and Intelligent Informatics.

This study focuses on video distribution services that many people use on a daily basis, and analyzes how viewers' feelings for promotional videos (PVs) are linked to their intent. In particular, by comparing viewers in Japan and Singapore, we clarified the impact of cultural backgrounds on how we perceive emotions.

Points of Research

1.   Type of viewer by AI

Using fuzzy c-means clustering, one of the AI technologies, we analyzed the viewer's complex emotional response to PV. As a result, it was found that viewers can be classified into four groups with different characteristics of "high evaluation", "medium evaluation", "mixed evaluation", and "low evaluation". With this analysis, it is possible to identify what kind of emotions people are more likely to lead to viewing.

2.   The key to watching is “joy” and “excitement”

Statistical analysis reveals that the two positive emotions that viewers feel, Joy and Excitement, are extremely strongly influenced by their subsequent desire to see. In particular, the effect of "joy" was significant, showing that these emotions could account for about 54% of the variation in intent of viewing.

3.   Discover cultural differences in how emotions are perceived

When we compared responses from 71 Japanese students and 27 Singapore students, we found cultural differences in emotional responses. In particular, in the "joy" section, students in Singapore showed a statistically significant higher rating than Japanese students (p < 0.05), indicating the importance of promotion taking into account cultural backgrounds.

4.   Significance of Research and Future Prospects

The results of this study provide practical suggestions for creating a more effective video promotion strategy. For example, by creating images that strategically bring out emotions such as "joy" and "excitement" according to the target country and cultural sphere, it is possible to maximize the motivation to watch.

In the future, we will conduct surveys using videos of more diverse countries, age groups, and genres, and aim to gain more versatile knowledge. We will continue to promote research that solves various issues in human society through the power of data science.

Comments from the author

“YouTube and Netflix are now essential to our lives. In this context, how does the appeal of the work that the creator wants to deliver resonate with the viewer? We started this research with the desire to elucidate the mechanism by the power of data science. We hope that this achievement will help create content that moves people's hearts across borders. Anyone can read this paper for free, so please take a look at it.

Paper information

• Title of the paper: An Analysis of Viewing Intentions for Promotional Videos Using Fuzzy c-Means Clustering: A Comparative Study Between Japan and Singapore
  (Japanese translation: Analysis of the intent of viewing promotional videos using fuzzy c-means clustering: Comparative study between Japan and Singapore)
• Author: Naruki Shirahama, Naofumi Nakaya, Kenji Moriya, Kazuhiro Koshi, Keiji Matsumoto, and Satoshi Watanabe
• Publications: Journal of Advanced Computational Intelligence and Intelligent Informatics, Vol. 29, No.4, pp.847-856
• Date of publication: July 20, 2025
• DOI (link to paper): https://doi.org/10.20965/jaciii.2025.p0847

This paper is published in JACIII, an international English journal in the field of AI and computational intelligence, and is also included in the world's largest abstract and citation database, Scopus. In addition, the Journal Impact Factor, which is one of the indicators of the degree of impact of academic journals, is rated at 0.8 (2024). This paper is published in open access (CC BY-ND 4.0), and anyone can read the full text for free.

[Contact information]
Shimonoseki City University
Corporate Planning Department Public Relations and Branding Strategy Division
TEL. 083-253-8967