Paper 1
■書誌情報
Sunil Kumar Maurya, Makoto Kawano, Yusuke Iwasawa, Yutaka Matsuo: Temporal-Aware Graph Attention for Sequential Knowledge Tracing, Proceedings of the 35th ACM International Conference on Information and Knowledge Management (CIKM 2026), November 2026
■概要
Knowledge tracing aims to model student learning trajectories and predict future performance from historical question–response interactions. Many recent works approach knowledge tracing with graph-based modeling, where concept dependencies (question-concept relations) are represented as graphs and learning is driven by message passing or attention over these structures. Existing graph-based KT methods often rely on predefined or weakly learned graphs and lack a principled mechanism to jointly account for temporal distance and contextual relevance when aggregating past interactions. As a result, they may overemphasize non-informative signals while under-utilizing informative recent context. In this work, we introduce a graph attention-based knowledge tracing model, TAGAKT, and reframe the problem as implicit graph learning over a causal directed acyclic graph. Our proposed graph attention model incorporates context-aware, learnable temporal bias that adjusts the importance of historical interactions based on their distance to the current prediction point. In addition, we integrate an efficient channel recalibration module that performs lightweight cross-channel reweighting to suppress noisy feature dimensions arising from interaction aggregation, improving prediction robustness. Extensive experiments on six datasets show that TAGAKT outperforms baselines on most datasets and achieves the best overall rank.
Paper 2
■書誌情報
Justin Dachille*, Aurora Rossi*, Sunil Kumar Maurya, Frederik Mallmann-Trenn, Xin Liu, Frédéric Giroire, Tsuyoshi Murata, Emanuele Natale: Degree-Mass Message Passing for Betweenness Ranking in Directed and Undirected Networks, Proceedings of the 35th ACM International Conference on Information and Knowledge Management (CIKM 2026), November 2026 (*equal contribution)
■概要
Computing the importance of nodes in networks is a long-standing fundamental problem that has driven extensive study of various centrality measures. A particularly well-known centrality measure is betweenness centrality, whose exact computation becomes prohibitive on large-scale networks. Graph Neural Network (GNN) models have thus been proposed to predict the ranking of nodes by betweenness centrality. However, existing GNN-based methods either have graph-size-dependent parameter counts or are limited to undirected graphs. We propose a lightweight GNN architecture that exploits the empirically observed relationship between betweenness centrality and multi-hop degree mass. This motivates the use of degree masses as size-invariant node features. To improve generalization, we train on synthetic graphs whose degree distributions more closely match those of real-world networks, including directed and undirected scale-free graphs and uniformly directed hyperbolic random graphs. We evaluate our model on 14 real-world networks spanning eight domains, including social, email, and citation networks, across both directed and undirected regimes. The experiments show that our model improves the Kendall τb correlation up to 24.6% on undirected and 10.9% on directed graphs, while using 56x fewer parameters than the lightest competing GNN baseline and achieving competitive inference time, with up to a 24.5x speedup on selected directed graphs.
