๐ค AI Summary
To address the challenge of effectively identifying and interpreting long-term player strategies from high-dimensional, noisy action-sequence data in educational games, this paper proposes a kinetic visualization framework integrating dynamic visual encoding with layered hybrid rendering. The method models player behavior trajectories using parametric interpolation curves and constructs interpretable data narratives through multi-layer semantic fusion and temporally continuous animation encoding, significantly mitigating overplottingโa common issue in traditional visualizations. By jointly enforcing structural constraints and capturing behavioral dynamics, the framework preserves original sequential state information while enhancing the readability and traceability of strategic patterns. Evaluated in authentic classroom settings, the tool enabled domain experts to uncover latent learning pathways and characterize strategy evolution, thereby demonstrating its effectiveness and practical utility for explainable analysis of educational interaction data.
๐ Abstract
Game-Based Learning has proven to be an effective method for enhancing engagement with educational material. However, gaining a deeper understanding of player strategies remains challenging. Sequential game-state and action-based tracking tools often gather extensive data that can be difficult to interpret as long-term strategy. This data presents unique problems to visualization, as it can be fairly natural, noisy data but is constrained within synthetic, controlled environments, leading to issues such as overplotting which can make interpretation complicated. We propose an animated visual encoding tool that utilizes kinetic visualization to address these issues. This tool enables researchers to construct animated data narratives through the configuration of parameter interpolation curves and blending layers. Finally, we demonstrate the usefulness of the tool while addressing specific interests as outlined by a domain expert collaborator.