Boosting Team Modeling through Tempo-Relational Representation Learning

TL;DR

This paper introduces a novel tempo-relational neural architecture for team modeling that captures temporal interactions and predicts multiple team constructs, outperforming existing methods and providing interpretable insights for real-world applications.

Contribution

The paper proposes a new neural architecture that models temporal and relational aspects of team dynamics and extends it for multi-task learning, improving prediction accuracy and efficiency.

Findings

Outperforms temporal-only and relational-only models in team performance prediction.

Multi-task extension reduces training and inference time significantly.

Provides interpretable insights and actionable recommendations for team improvement.

Abstract

Team modeling remains a fundamental challenge at the intersection of Artificial Intelligence and Social Sciences. Although a variety of computational models have been proposed in the last two decades, most fail to integrate Social Sciences insights, such as the critical role of temporal interactions in shaping team dynamics, and do not meet key practical requirements for real-world applications, including the ability to provide real-time, actionable recommendations to enhance team performance. To address these limitations, in this paper, we propose a novel tempo-relational neural architecture that jointly models interactions between team members and the evolution of team dynamics through temporal graphs. We additionally propose a multi-task extension of the architecture that learns shared social embeddings for team members enabling the simultaneous prediction of multiple team constructs (e.g., Emergent Leadership, Leadership Style, and Teamwork components). Experiments on two state-of-the-art team datasets show that our tempo-relational architecture out performs temporal-only and relational-only approaches for team performance prediction, and that its multi-task extension substantially reduces training and inference time without loss of predictive performance. Finally, the integration of explainability techniques within the proposed architectures provides interpretable insights and actionable recommendations to support team improvement. These strengths make our approach particularly well-suited for human-centered artificial intelligence applications, such as intelligent decision-support systems in high-stakes collaborative environments.