Bibliometric Analysis of Artificial Intelligence in Team Sports: Research Trends, Collaborations, and Creative Insights

Abstract

This paper presents a bibliometric analysis of Scopus-indexed publications from 2000 to 2024 to examine the development of deep learning in team sports, particularly in real-time player tracking, performance analytics, and wearable-based monitoring of athletes. The publication trends, collaboration networks, and thematic structures were explored using Bibliometrix (R) and VOSviewer based on the co-occurrence of keywords, co-citation, and clustering. The findings indicate that after 2016, research production has increased significantly, with convolutional neural networks, temporal models, and new transformer architectures emerging as prominent approaches. The analysis identified three predominant research streams: computer vision–based detection and tracking, machine learning–driven performance analytics, and wearable-sensor applications. Object detectors and multi-object tracking frameworks used together to create a computer vision pipeline are now the backbone of real-time sports analytics, and predictive models are increasingly used to support tactical decision-making and workload evaluation. In spite of these developments, there are still issues concerning model interpretability, resiliency to dynamic environments, data privacy, and incorporation into the coaching workflows. It is recommended that future studies focus on multimodal systems combining visual, temporal, and physiological cues, as well as explainable artificial intelligence methods to enhance transparency and trust among practitioners. Further, this paper describes a system-based AI architecture that integrates vision-based analysis, time-based learning, and wearable sensing to support data-driven decision-making. Overall, this bibliometric mapping offers a systematic understanding of the new research areas and sets the stage toward building operational, reliable, and human-centric AI systems in team sports.


Received: 28 November 2025 | Revised: 15 April 2026 | Accepted: 25 May 2026



Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.



Data Availability Statement

Data are available from the corresponding author upon reasonable request.



Author Contribution Statement

Doha Lefhal: Conceptualization, Methodology, Software, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization, Project administration, Funding acquisition. Ali Ouacha: Validation, Supervision. Abdeslam El Harraj: Resources. Soumia Ziti: Supervision.