Article: Stress level identification during sports competitions using clustering and recommendation algorithms Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.89 pp.39 - 62 Abstract: In competitive sports, sources of stress for athletes - such as critical moments in competition, audience distractions, and coaches' expectations - vary greatly from person to person. Traditional questionnaire-based methods are subjective and lag behind events, making it difficult to support real-time interventions. Single recommendation algorithms, lacking prior knowledge of stressor types, tend to produce 'one-size-fits-all' strategies, which can exacerbate mental fatigue. To address this challenge, this paper proposes an identification framework that integrates hierarchical clustering with hybrid collaborative filtering: first, typical stressor patterns are extracted through clustering, and then personalised stress relief strategies are generated based on these patterns. Validation using publicly available physiological stress datasets shows that, compared to traditional collaborative filtering, our method improves accuracy from 78.3% to 86.7%, increases the area under the curve from 0.83 to 0.89, and raises the normalised discounted cumulative gain @10 from 0.72 to 0.81. Inderscience Publishers - linking academia, business and industry through research

Title: Stress level identification during sports competitions using clustering and recommendation algorithms

Authors: Tong Zhang; Xiangke Yang

Addresses: Department of Sports, Henan University of Animal Husbandry and Economy, Zhengzhou, 480018, China ' College of Food and Biological Engineering, Henan University of Animal Husbandry and Economy, Zhengzhou, 480018, China

Abstract: In competitive sports, sources of stress for athletes - such as critical moments in competition, audience distractions, and coaches' expectations - vary greatly from person to person. Traditional questionnaire-based methods are subjective and lag behind events, making it difficult to support real-time interventions. Single recommendation algorithms, lacking prior knowledge of stressor types, tend to produce 'one-size-fits-all' strategies, which can exacerbate mental fatigue. To address this challenge, this paper proposes an identification framework that integrates hierarchical clustering with hybrid collaborative filtering: first, typical stressor patterns are extracted through clustering, and then personalised stress relief strategies are generated based on these patterns. Validation using publicly available physiological stress datasets shows that, compared to traditional collaborative filtering, our method improves accuracy from 78.3% to 86.7%, increases the area under the curve from 0.83 to 0.89, and raises the normalised discounted cumulative gain @10 from 0.72 to 0.81.

Keywords: identification of competitive stressors; cluster analysis; collaborative filtering; personalised recommendations.

DOI: 10.1504/IJICT.2026.155693

International Journal of Information and Communication Technology, 2026 Vol.27 No.89, pp.39 - 62

Received: 09 Apr 2026
Accepted: 26 May 2026
Published online: 10 Aug 2026 *

Title: Stress level identification during sports competitions using clustering and recommendation algorithms

Keep up-to-date