Data for: A Knowledge-Driven Framework for Modeling Dynamic Social Networks via Temporal Knowledge Graphs

This dataset contains all numerical values supporting the results reported in the paper "A Knowledge-Driven Framework for Modeling Dynamic Social Networks via Temporal Knowledge Graphs". Included are: (1) link prediction MRR values by relation type for all methods (Table 1); (2) ablation study results (Table 2); (3) knowledge graph construction statistics including entity counts, relation type distributions, and temporal coverage; (4) community modularity values; (5) model hyperparameters. The historical data was derived from the China Biographical Database (CBDB), Harvard University.