Identifying communal user needs for quality of experience (QoE) evaluation and improvement focused on user-centered design.

Abstract

The rise of intelligent systems has led to the emergence of new types of interactions and diverse purposes of technology use. As a result, the importance of Quality of Experience (QoE) is becoming increasingly important in fulfilling user needs and ensuring high satisfaction. However, current measurement and improvement methods of QoE are primarily based on objective QoS metrics rather than the highly subjective nature of user needs. Additionally, evaluation models are developed separately for each technological domain, focusing just on domain-specific parameters, leading to inconsistencies when applied to other domains. To address this, the present study selected ten technological domains and employed BERTopic and thematic analysis to identify user needs. By identifying patterns in shared aspects, the study defined eleven common dimensions referred to as "communal user needs." To verify the reliability of these communal user needs, three researchers independently evaluated 1000 reviews to determine which communal user need each review corresponded to. The resulting Krippendorff's α was 0.805, indicating a sufficient level of inter-rater reliability. The findings suggest that, while user needs may be expressed differently at the individual technology level, they align toward communal user needs at an abstract level. Furthermore, the variation in the number of times the user needs appeared across different technological domains in a similar pattern suggests that a more universal QoE model could be constructed. This could be done by incorporating the dynamic nature of user needs through relative importance adjustments. The proposed methodology of identifying communal user needs for QoE measurement is expected to serve as a reliable evaluation tool that can effectively respond to the rapidly evolving technology market.