Bandung, November 2025 — Researchers from the Research Center for Information and Communication Technology (PPTIK), Institut Teknologi Bandung (ITB), have published a study in IEEE Access (DOI: 10.1109/ACCESS.2025.3635242), titled “A Study on the Correlation Between User Sentiments and QoE Parameters in Video Streaming Services.” The research presents a data-driven approach to understanding video streaming quality through user perceptions expressed on social media.
Published on November 20, 2025, the study was conducted by researchers from PPTIK and the School of Electrical Engineering and Informatics (STEI) ITB, in collaboration with Pusan National University, South Korea. The research involved Fahmi Candra Permana, Fetty Fitriyanti Lubis, Armein Z. R. Langi, Aditya Pradana, Andriyan Bayu Suksmono, Kusprasapta Mutijarsa, Dion Tanjung, Yoanes Bandung, and Jong-Deok Kim.
Rethinking Video Streaming Quality Evaluation
Traditional video streaming quality assessments often emphasize technical indicators such as bandwidth, latency, and video resolution. However, these measurements do not fully capture how users experience digital services.
Quality of Experience (QoE) provides a broader perspective by incorporating user expectations, content preferences, viewing contexts, and system performance. This evolution reflects a shift from technology-centered evaluation toward a more comprehensive, user-centered approach.
Figure 1. Conceptual evolution of QoE assessment, from technology-centric measurements to interactive, user-centered ecosystems. Source: Figure 1 of the published article.
Building on this perspective, the research explores social media feedback as a complementary approach to conventional QoE assessments, which often require time-consuming surveys and controlled experiments.
Analyzing User Perceptions Across Five Streaming Platforms
The study examined five major video streaming services, comprising four international platforms and one Indonesian service. Using data collected from X (formerly Twitter), researchers initially obtained 11,280 user posts, subsequently filtered into 2,556 relevant comments for analysis.
The methodology combined text mining, sentiment analysis, and social network analysis to examine user perceptions. Comments were classified into four principal QoE dimensions: Human, Content, Context, and System, with an additional Others category for unrelated feedback. Three domain experts conducted the annotation, achieving a Krippendorff’s Alpha coefficient of 0.76, indicating substantial inter-annotator agreement.
Key Findings: Content and Human Factors Dominate
The findings reveal that 98.16% of analyzed comments were associated with established QoE parameters, demonstrating the relevance of social media feedback for understanding video streaming experiences.
Content emerged as the most frequently discussed factor, accounting for an average of 40.69% of categorized user feedback across platforms, followed by Human factors at 28.27%. These findings highlight the importance of content availability, personal preferences, and emotional engagement in shaping user perceptions.
Figure 2. Comparative analysis of QoE parameters across five video streaming platforms, highlighting differences in content, human, context, and system-related factors. Source: Figure 10 of the published article.
The analysis also identified platform-specific concerns, including live sports streaming reliability, regional content availability, subscription accessibility, and application stability. These findings suggest that service providers should balance content strategies with technical improvements and user-centered service development.
Figure 3. Classification of social media feedback into QoE and Non-QoE parameters across five video streaming platforms. Source: Figure 11 of the published article.
Toward More User-Centered Multimedia Services
This research demonstrates how social media analytics can complement conventional QoE assessments by providing insights into real-world user perceptions. The findings establish a foundation for developing more scalable, adaptive, and data-driven approaches to multimedia service evaluation.
Through this international research collaboration, PPTIK ITB continues to contribute to advancing multimedia communication research and supporting the development of digital services that better reflect user expectations and experiences.
Publication: IEEE Access, Volume 13, 2025, pp. 198659–198676.
