CognTKE: A Cognitive Temporal Knowledge Extrapolation Framework

TL;DR

CognTKE tackles the challenge of extrapolating future facts on temporal knowledge graphs by integrating cognitive-inspired reasoning with graph-based evidence. It introduces a Temporal Cognitive Relation Digraph (TCR-Digraph) that fuses global, one-hop historical relations with local, multi-hop histories, and employs a two-stage reasoning process: a global shallow (System 1) and a local deep (System 2) engine realized via TR-Component and TR-GAT, respectively. A formal theorem links CognTKE’s attention-driven subgraphs to temporal logical rules, supporting interpretability while maintaining strong predictive performance. Experiments on four benchmarks demonstrate gains over state-of-the-art baselines and robust zero-shot reasoning, with code available for reproducibility and further exploration of interpretability in TKG extrapolation.

Abstract

Reasoning future unknowable facts on temporal knowledge graphs (TKGs) is a challenging task, holding significant academic and practical values for various fields. Existing studies exploring explainable reasoning concentrate on modeling comprehensible temporal paths relevant to the query. Yet, these path-based methods primarily focus on local temporal paths appearing in recent times, failing to capture the complex temporal paths in TKG and resulting in the loss of longer historical relations related to the query. Motivated by the Dual Process Theory in cognitive science, we propose a \textbf{Cogn}itive \textbf{T}emporal \textbf{K}nowledge \textbf{E}xtrapolation framework (CognTKE), which introduces a novel temporal cognitive relation directed graph (TCR-Digraph) and performs interpretable global shallow reasoning and local deep reasoning over the TCR-Digraph. Specifically, the proposed TCR-Digraph is constituted by retrieving significant local and global historical temporal relation paths associated with the query. In addition, CognTKE presents the global shallow reasoner and the local deep reasoner to perform global one-hop temporal relation reasoning (System 1) and local complex multi-hop path reasoning (System 2) over the TCR-Digraph, respectively. The experimental results on four benchmark datasets demonstrate that CognTKE achieves significant improvement in accuracy compared to the state-of-the-art baselines and delivers excellent zero-shot reasoning ability. \textit{The code is available at https://github.com/WeiChen3690/CognTKE}.

Figures (9)

• Figure 1: An example illustrates longer history temporal relation path loss. Relations are denoted using solid lines with different colors, and inverse relations are denoted using dashed lines. Black solid line denotes the identity relation.
• Figure 2: An illustrative diagram of the proposed CognTKE architecture.
• Figure 3: An illustrative of TCR-Digrph $\hat{\mathcal{G}}_{e_1,e_3|3}$ formed by the TKG.
• Figure 4: Visualization of the learned structures. Dashed lines mean inverse relations. entities are indicated by the blue rectangles.
• Figure 5: The Component of QTR_GRU.