Key Nodes Evaluation for Human Proximity Networks Based on Gravity Model
摘要
Human proximity networks describe the offline interactions among individuals. Evaluating key nodes in networks has research value. We propose a key nodes evaluation method for human proximity networks based on gravity model. Sling the networks to obtain the sequence of network snapshots, the relative entropy and node interaction frequency are used to certain the slice length. Node aggregation degree, effective distance, and node pair similarity are used to construct evaluation indicators. Gravity model is used to calculate node importance to evaluate key nodes. Results on three real-word datasets show our method achieved better performance.