Identifying the influential nodes is one of the research focuses in network information mining. Many centrality measures used to evaluate influence abilities of nodes can’t balance between high accuracy and low time complexity. The NL centrality based on the neighbors and importance of edges is proposed which considers the second-degree neighbor’s impact on the influence of a node and utilizes the connectivity and unsubstitutability of edge to distinguish topological position of a node. In order to evaluate the accuracy of NL centrality, the SIR model is used to simulate the process of virus propagation in four real-world networks. Experiment results of monotonicity, validity and efficiency demonstrate that the NL centrality has a competitive performance in distinguishing the influence of nodes and it is suitable for large-scale networks because of the high efficiency in computation.
23 March 2023 in ACM Transactions on Asian and Low-Resource Language Information Processing
Chinese Named Entity Recognition (NER) is an essential task in natural language processing, and its performance directly impacts the downstream tasks. The main challenges in Chinese NER are the high dependence of named entities on context and the lack of word boundary information. Therefore, how to integrate relevant knowledge into the corresponding entity has become the primary task for Chinese NER. Both the lattice LSTM model and the WC-LSTM model did not make excellent use of contextual information. Additionally, the lattice LSTM model had a complex structure and did not exploit the word information well. To address the preceding problems, we propose a Chinese NER method based on the deep neural network with multiple ways of embedding fusion. First, we use a convolutional neural network to combine the contextual information of the input sequence and apply a self-attention mechanism to integrate lexicon knowledge, compensating for the lack of word boundaries. The word feature, context feature, bigram feature, and bigram context feature are obtained for each character. Second, four different features are used to fuse information at the embedding layer. As a result, four different word embeddings are obtained through cascading. Last, the fused feature information is input to the encoding and decoding layer. Experiments on three datasets show that our model can effectively improve the performance of Chinese NER.
Traffic control of urban road networks during emergency rescues is conducive to rapid rescue in the affected areas. However, excessive control will lead to negative impacts on the normal traffic order. We propose a novel model to optimize the traffic control scheme during the post-disaster emergency rescue period named PD-TCM (post-disaster traffic control model). In this model, the vertex and edge betweenness indexes of urban road networks are introduced to evaluate the controllability of the road sections. The gravity field model is also used to adjust the travel time function of different road sections in the control and diverging domains. Experimental results demonstrate that the proposed model can obtain the optimal traffic control scheme efficiently, which gives it the ability to meet the demand of emergency rescues as well as reducing the disturbances caused by controls.
The article “Identifying influential nodes in complex networks based on Neighbours and edges”, written by Zengzhen Shao, Shulei Liu, Yanyu Zhao, and Yanxiu Liu, was originally published...