Network Structure Characteristics Research Articles

Critical station identification and robustness analysis of urban rail transit networks based on comprehensive vote-rank algorithm

Urban rail transit networks have become the most important traffic modal and greatly improve the traffic environment and travel efficiency of metropolis, and the critical station identification and robustness assessment of urban rail transit networks have become more and more important. In this paper, we propose a comprehensive vote-rank (CVR) algorithm to identify the critical stations, and adopt mutual information and k-center clustering algorithm to analyze the similar characteristics of passenger flow on different dates of urban rail transit networks. Meanwhile, Shanghai metro network is chosen as the example and we find that the optimal classification of passenger flow on different dates is 2 categories, and the passenger flows have the significant differences between conventional day and unconventional day. The results show that the critical stations on unconventional day are farther from the city center than the critical stations on conventional day, and some critical stations locate in the suburbs on unconventional day. Moreover, we study the structural and functional robustness characteristics of urban rail transit networks subjected to continuous malicious attacks to the top critical nodes, and the results show that Shanghai metro network has the worse robustness subjected to malicious attacks. Compared with the existed methods, we can declare that the proposed CVR algorithm is the most effective method for critical station identification and robustness assessment among the discussed methods in this paper, and this study will provide theoretical and practical support for the safe operation of urban rail transit networks.

TAMNR: a network embedding learning algorithm using text attention mechanism.

Because many existing algorithms are mainly trained based on the structural features of the networks, the results are more inclined to the structural commonality of the networks. These algorithms ignore the rich external information and node attributes (such as node text content, community and labels, etc.) that have important implications for network data analysis tasks. Existing network embedding algorithms considering text features usually regard the co-occurrence words in the node's text, or use an induced matrix completion algorithm to factorize the text feature matrix or the network structure feature matrix. Although this kind of algorithm can greatly improve the network embedding performance, they ignore the contribution rate of different co-occurrence words in the node's text. This article proposes a network embedding learning algorithm combining network structure and co-occurrence word features, also incorporating an attention mechanism to model the weight information of the co-occurrence words in the model. This mechanism filters out unimportant words and focuses on important words for learning and training tasks, fully considering the impact of the different co-occurrence words to the model. The proposed network representation algorithm is tested on three open datasets, and the experimental results demonstrate its strong advantages in node classification, visualization analysis, and case analysis tasks.

A data-driven approach shows that individuals' characteristics are more important than their networks in predicting fertility preferences.

People's networks are considered key in explaining fertility outcomes-whether people want and have children. Existing research on social influences on fertility is limited because data often come from small networks or highly selective samples, only few network variables are considered, and the strength of network effects is not properly assessed. We use data from a representative sample of Dutch women reporting on over 18 000 relationships. A data-driven approach including many network characteristics accounted for 0 to 40% of the out-of-sample variation in different outcomes related to fertility preferences. Individual characteristics were more important for all outcomes than network variables. Network composition was also important, particularly those people in the network desiring children or those choosing to be childfree. Structural network characteristics, which feature prominently in social influence theories and are based on the relations between people in the networks, hardly mattered. We discuss to what extent our results provide support for different mechanisms of social influence, and the advantages and disadvantages of our data-driven approach in comparison to traditional approaches.

Predicting circRNA-miRNA interactions utilizing transformer-based RNA sequential learning and high-order proximity preserved embedding

A key regulatory mechanism involves circular RNA (circRNA) acting as a sponge to modulate microRNA (miRNA), and thus, studying their interaction has significant medical implications. In this field, there are currently two pressing issues that remain unresolved. Firstly, due to the scarcity of verified interactions, we require a minimal amount of samples for training. Secondly, the current models lack interpretability. Therefore, we propose SPBCMI, a method that combines sequence features extracted using the Bidirectional Encoder Representations from Transformer (BERT) model and structural features of biological molecule networks extracted through graph embedding to train a GBDT (Gradient-boosted decision trees) classifier for prediction. Our method yielded an AUC of 0.9143, which is currently the best for this problem. Furthermore, in the case study, SPBCMI accurately predicted 7 out of 10 circRNA-miRNA interactions. These results show that our method provides an innovative and high-performing approach to understanding the interaction between circRNA and miRNA.

Research on the influence of chain shareholder network on enterprise green innovation

Green innovation plays a pivotal role in the high-quality development of social economy, but its high risk and investment can daunt enterprises. As a result, business stakeholders often use interpersonal relationships to start to overcome boundaries to obtain and disseminate green innovation information and resources. Based on the number of green innovation patent applications from 2011 to 2020, this paper constructs an undirected weighted network of enterprise chain shareholders and empirically examines the influence of centrality and structural hole characteristics of social networks on green innovation. Additionally, a heterogeneity analysis is conducted from the perspectives of core-periphery structure and CEO overseas experience. The findings are beneficial for enterprises to have achieve a more comprehensive understanding of the role of their positional advantages in the network on green innovation and provide new ideas and directions for firms to improve their green innovation level.

LGNN: a novel linear graph neural network algorithm.

The emergence of deep learning has not only brought great changes in the field of image recognition, but also achieved excellent node classification performance in graph neural networks. However, the existing graph neural network framework often uses methods based on spatial domain or spectral domain to capture network structure features. This process captures the local structural characteristics of graph data, and the convolution process has a large amount of calculation. It is necessary to use multi-channel or deep neural network structure to achieve the goal of modeling the high-order structural characteristics of the network. Therefore, this paper proposes a linear graph neural network framework [Linear Graph Neural Network (LGNN)] with superior performance. The model first preprocesses the input graph, and uses symmetric normalization and feature normalization to remove deviations in the structure and features. Then, by designing a high-order adjacency matrix propagation mechanism, LGNN enables nodes to iteratively aggregate and learn the feature information of high-order neighbors. After obtaining the node representation of the network structure, LGNN uses a simple linear mapping to maintain computational efficiency and obtain the final node representation. The experimental results show that the performance of the LGNN algorithm in some tasks is slightly worse than that of the existing mainstream graph neural network algorithms, but it shows or exceeds the machine learning performance of the existing algorithms in most graph neural network performance evaluation tasks, especially on sparse networks.

How ICT capital affects the spatial correlation of energy consumption-a new perspective based on spatially correlation network.

ICT has significantly transformed the traditional energy production and service methods, changed the correlation characteristics of energy consumption network, and contributed to the establishment of cross-regional, open, and synergistic energy ecological networks. In this paper, social network analysis (SNA), dynamic SYS-GMM model, and mediating effects model are employed to deliberate on the mechanism of ICT capital's influence on the spatial correlation of energy consumption from 2000 to 2019. Firstly, this study employs an enhanced gravity model to precisely delineate the spatial correlation network of energy consumption in China, further applies the SNA to analyze the network structural characteristics, and then uses the econometric model to investigate the influence mechanism and heterogeneity of ICT capital on the spatial correlation of energy consumption. The study demonstrates a progressive spatial correlation in energy consumption in China, with eastern provinces emerge as the center of the network, assuming the position of the "dominant player." Conversely, the central provinces act as the "bridge," and western provinces are positioned at the periphery, referred to as the "edge" of the network. ICT capital contributes to improving the energy consumption spatial correlation, mainly by stimulating green technology innovation, promoting industrial structure optimization, accelerating human capital accumulation, and reducing income inequality. As ICT capital expands, the eastern region becomes more preeminent as a network hub for energy consumption, the central region increases its dominance slightly, and the western region's marginal position does not change significantly. Furthermore, the presence of ICT capital significantly enhances the intensity of energy consumption spatial correlation more prominently for low-carbon pilot areas and high Internet level areas. This study guides provinces to fully utilize ICT capital to reach collaborative energy-saving goals, and promotes the breaking down of regional competitive barriers in energy systems to build cooperative energy conservation ecological networks.

Spatial Correlation Network Structure of and Factors Influencing Technological Progress in Citrus-Producing Regions in China

In this study, the transcendental logarithmic cost function model was used to measure the rate of technological progress in seven major mandarin-producing regions and seven major tangerine-producing regions in China from 2006 to 2021. The modified gravity model was used to establish spatial correlation networks. The social network analysis method was used to analyze the characteristics of the overall network structure and the individual network structure of the spatial correlation networks of citrus-production technology progress, and the quadratic assignment procedure was used to analyze the factors influencing the spatial network. The results show the production of Chinese mandarins and tangerines is in the stage of technological progress in general, but the rate of progress is slowing down gradually, and the rate of mandarin-production technology progress is higher than that of tangerine-production technology progress. In terms of the overall network structure characteristics, the spatial networks of technological progress related to Chinese mandarin and tangerine production are becoming increasingly dense and complex, with obvious spatial spillover effects, but the network structure is relatively loose, and the polarization of the tangerine network is more serious. In terms of individual network structure characteristics, the relatively economically developed eastern regions have a higher status in terms of the spatial correlation network and a stronger role in controlling and dominating the resource elements needed for citrus-production technology progress. Education, informatization, economic development, innovation support, and financial support are important factors influencing the formation of the spatial association network of citrus-production technology progress in China.

Social drug dealing: how peer-to-peer fintech platforms have transformed illicit drug markets

Digital platforms have revolutionized the way illegal drug trafficking is taking place. Modern drug dealers use social network platforms, such as Instagram and TikTok, as direct-to-consumer marketing tools. But apart from the marketing side, drug dealers also use fintech payment apps to engage in financial transactions with their clients. In this work, we leverage a large dataset from Venmo to investigate the digital money trail of drug dealers and the social networks they create. Using text and social network analytics, we identify two types of illicit users: mixed-activity participants and heavy drug traffickers and build a random forest classifier that accurately predicts both types of illicit nodes. We then investigate the social network structure of drug dealers on Venmo and find that heavy drug traffickers share similar network characteristics with previous literature findings on drug trafficking networks. However, mixed-activity participants exhibit different patterns of network structure characteristics, including a higher clustering coefficient, suggesting that they may be accessing multiple networks and bridging those networks through their illicit activities. Our findings highlight the importance of distinguishing between these two types of illicit users and provide law enforcement agencies with valuable insights that can aid in combating illegal drug transactions in digital payment apps.

Rural development models in China: The participant’s perspective

This study aims to explore the feasible and sustainable rural development model in China from the perspective of participants. The paper proposed the concept of ‘the community of rural development’: the actors who participate in rural planning, housing and infrastructure construction, industrial development, environmental improvement and social progress, and analysed its evolutionary process from traditional to modern and social network structure characteristics. On this basis, the study proposed six models of rural development in China and selected six typical cases for detailed analysis. These models include village committee-led, peasant organization-led, township enterprise-led, government-led, external enterprise-led and NGO-led. Based on the methodology of social network analysis, the paper designed a simulated routing to analyse the network evolution characteristics of these models and elaborated four output indicators: clustering coefficient, average path length, density and degree centrality of the core actor. The results showed that the internal-actor-led model is superior to the external-actor-led model due to the close inter-relationships among the participants and the strong influence of the core actor. Finally, the paper proposed building partnerships in the community of rural development that are stable and harmonious over the long term and could promote rural development and sustainable renewal in a comprehensive way.

Integration of migration and attention flow data to reveal association of virtual–real dual intercity network structure

The study of intercity network structure from the perspective of flow space can effectively reveal the connection patterns, hierarchical laws, and spatial organization of cities. However, current studies mostly focus on a single real or virtual space, neglecting the comprehensive perspective. Furthermore, the research subject of network relationships is usually the complex original aligned network rather than the spatially cohesive network structure. In this study, spatial network structures are constructed and extracted based on the migration flow and attention flow, which represent people's objective behavior and subjective will, respectively. This study then reveals the association between virtual and real networks from three aspects: node location, node attributes, and local network structure characteristics. Results indicate that compared with the real network structure, the virtual network structure has a strong same-level node agglomeration and a significant node size difference. Spatial dislocations are observed in the high association of the various measures of China's real and virtual network structures. The degree of differences in network structures between virtual and real spaces varies spatially. This paper reveals the dynamics of urban development in different spaces, helping build a human-centered, high-quality space governance system for virtual and real spaces.

Efficient real-time detection of electrical equipment images using a lightweight detector model

Infrared technology holds significant importance in the detection of electrical equipment, as it has the capability to swiftly and securely identify electrical apparatus. To simplify the implementation of proficient detection frameworks for electrical equipment within constrained settings (like embedded apparatus), this study presents an enhanced, lightweight model of the single-shot multibox detector (SSD). This model specifically addresses the detection of multiple equipment objects within infrared imagery. The model realized the lightweight of the model by using the network structure characteristics of squeezenet to modify the backbone network of SSD, and compensated for the impact of the lightweight model on the detection accuracy by adding multiple convolutional layers and connecting branches to enhance the propagation ability and extraction ability of features. To ensure a comprehensive evaluation of the model’s detection capabilities, all the models discussed in this study employed the technique of random weight initialization. This approach was utilized to validate the optimal structure of the model and its performance. The experimentation was conducted on both the PASCAL VOC 2007 benchmark dataset and an infrared image dataset encompassing five distinct categories of electrical equipment found within substations. The experimental outcomes indicate that this model offers an efficient approach for achieving lightweight, real-time detection of electrical apparatus.

Examining the spatial networks of an urban agglomeration through the lens of multi-dimensional element flow: Evidence from western China

The Chengdu–Chongqing urban agglomeration (CCUA), as the only national-level city cluster in southwestern China, serves as a strategic support for the Yangtze River Economic Belt and an important demonstration area for promoting new urbanization in the country. The study of the networked characteristics of the CCUA contributes to a systematic understanding of its spatial connectivity patterns, optimization of spatial structure and layout, and promotion of high-quality regional development. In this study, we constructed models for traffic flow, information flow, migration flow, and composite flow to calculate the strength of connections between cities and the total flow of various elements in the CCUA. ArcGIS spatial visualization tools were used to depict the spatial connectivity patterns of the element flows within the CCUA. Additionally, social network analysis methods, including network density, centrality, and cohesive subgroups analysis, were employed to reveal the spatial network structure characteristics of the CCUA. The findings are as follows: (1) The overall structure of the cities within the CCUA is relatively loose, with significant differences in connectivity strength. It exhibits a west-strong and east-weak pattern, with Chengdu-Chongqing, Chengdu-Deyang, Chengdu-Mianyang, and Chengdu-Meishan occupying the top tier, while Zigong and Ya'an have relatively weak connections with other cities. Chengdu and Chongqing have prominent positions in the CCUA, with Chengdu having a more prominent core position compared to Chongqing, resulting in an overall hierarchical distribution of “1 + 1+7 + 7”. (2) The network density of the element flows in the CCUA is relatively low, indicating a generally weak element connectivity. The centrality of cities other than Chengdu and Chongqing is at a moderate to lower level, suggesting a weak overall resource connectivity capacity in the CCUA. (3) Most cities tend to form cohesive subgroups based on geographic proximity, while the cohesive subgroup in Chongqing is still in its early stages of development due to administrative boundaries. The research results quantitatively depict the spatial network structure characteristics of the CCUA, providing theoretical references for its development.

Measurement and analysis of the structural resilience of regional networks under the impact of COVID-19

The evaluation of the structural resilience of a regional network is of great significance for the improvement of the resilience of regions to cope with uncertain disturbances. In this study, COVID-19 was selected as a disturbance, and a regional network containing information, economic, population flow, and innovation links was constructed. Then, by combining the four dimensions of hierarchy, agglomeration, transmissibility, and diversity, the genetic algorithm was used to optimize the projection tracing model and the coupling coordination model to analyze the evolution of regional network resilience and its influencing factors in Zhejiang Province from 2019 to 2021. The results show that the network pattern and structural characteristics of the regional networks had different evolutions under the influence of COVID-19, among which the urban development path was found to be the most important influencing factor. Upgrading the corresponding city nodes from the two levels of network structure and function is an important way to optimize the regional network. In addition, the cities of regional networks can be divided into different types according to the development patterns. This paper provides a scientific basis and planning guidance for decision-making and the implementation of quantitative research on regional resilience at the provincial scale, as well as network optimization.

Brain structural covariance network features are robust markers of early heavy alcohol use.

Recently, we demonstrated that a distinct pattern of structural covariance networks (SCN) from magnetic resonance imaging (MRI)-derived measurements of brain cortical thickness characterized young adults with alcohol use disorder (AUD) and predicted current and future problematic drinking in adolescents relative to controls. Here, we establish the robustness and value of SCN for identifying heavy alcohol users in three additional independent studies. Cross-sectional and longitudinal studies using data from the Pediatric Imaging, Neurocognition and Genetics (PING) study (n = 400, age range = 14-22 years), the National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA) (n = 272, age range = 17-22 years) and the Human Connectome Project (HCP) (n = 375, age range = 22-37 years). Cases were defined based on heavy alcohol use patterns or former alcohol use disorder (AUD) diagnoses: 50, 68 and 61 cases were identified. Controls had none or low alcohol use or absence of AUD: 350, 204 and 314 controls were selected. Graph theory metrics of segregation and integration were used to summarize SCN. Mirroring our prior findings, and across the three data sets, cases had a lower clustering coefficient [area under the curve (AUC) = -0.029, P = 0.002], lower modularity (AUC = -0.14, P = 0.004), lower average shortest path length (AUC = -0.078, P = 0.017) and higher global efficiency (AUC = 0.007, P = 0.010). Local efficiency differences were marginal (AUC = -0.017, P = 0.052). That is, cases exhibited lower network segregation and higher integration, suggesting that adjacent nodes (i.e. brain regions) were less similar in thickness whereas spatially distant nodes were more similar. Structural covariance network (SCN) differences in the brain appear to constitute an early marker of heavy alcohol use in three new data sets and, more generally, demonstrate the utility of SCN-derived metrics to detect brain-related psychopathology.

Synergy between Electric Vehicle Manufacturers and Battery Recyclers through Technology and Innovation: A Game Theory Approach

Power battery recycling (PBR) has triggered profound changes in the industrial chain of electric vehicles (EVs). The PBR innovation network provides information channels and resource conditions for enterprises, but the mechanism of its impact on the synergistic innovation benefits and sustainable development ability of EV and PBR enterprises still needs further exploration. In this paper, we collect patent data for PBR from 2012 to 2020, identify the structural characteristics of innovation networks, and construct a synergy game model for PBR technology, aiming to analyze the synergistic effect of network embedding and knowledge spillover in PBR enterprises on technological innovation. First, we find that the PBR innovation network exhibits the small-world effect, which has a double-edged sword effect on technological cooperation innovation. Second, structural holes benefits of the main body of PBR technological innovation have a significant impact on cooperation innovation behavior. Third, the enhancement of the relevance and deep complementarity of knowledge cooperation is sufficient to make up for the input cost of PBR technological cooperation innovation, with additional benefits created by the increase in the output of structural holes. However, companies tend to be more inclined toward non-cooperative innovation as the knowledge spillover effect of the innovation network increases.

Structural Characteristics of Intergovernmental Water Pollution Control Cooperation Networks Using Social Network Analysis and GIS in Yangtze River Delta Urban Agglomeration, China

Water pollution exhibits distinct negative externalities, necessitating trans-regional collaborative governance among basin governments. However, few studies have examined the structural features of water pollution collaboration networks among local governments in China from a spatial analysis perspective. This study focuses on 27 central cities in the Yangtze River Delta, collecting 109 policy texts and evidence of cooperative actions on water pollution governance among these cities. By utilizing a combination of social network analysis and GIS spatial analysis, the research visualizes the results and delves into the overall structure and internal features of the network. The results indicate that the density of the water pollution cooperation network is 0.75, suggesting that a relatively stable and closely connected network for collaborative governance of water pollution has been formed. Furthermore, the water pollution cooperation in the Yangtze River Delta exhibits a typical “multi-center” network structure, with Shanghai–Suzhou, Nanjing, and Hangzhou as the core, forming three city clusters with tighter cooperation. Suzhou, Shanghai, and Jiaxing have the highest degree of centrality, which are 51, 46, and 44, respectively. The analysis of degree centrality reveals that cities with higher levels of economic development or those that serve as provincial capitals often play a leading role in the cooperation network. The study also observes that adjacent cities or local governments closer to the core cities are more likely to establish cooperative relationships; this phenomenon is not limited by provincial administrative boundaries.

Mathematical modeling and topological graph description of dominating David derived networks based on edge partitions

Chemical graph theory is a well-established discipline within chemistry that employs discrete mathematics to represent the physical and biological characteristics of chemical substances. In the realm of chemical compounds, graph theory-based topological indices are commonly employed to depict their geometric structure. The main aim of this paper is to investigate the degree-based topological indices of dominating David derived networks (DDDN) and assess their effectiveness. DDDNs are widely used in analyzing the structural and functional characteristics of complex networks in various fields such as biology, social sciences, and computer science. We considered the FN*, {M}_{2}^{*}, and {HM}_{N} topological indices for DDDNs. Our computations' findings provide a clear understanding of the topology of networks that have received limited study. These computed indices exhibit a high level of accuracy when applied to the investigation of QSPRs and QSARs, as they demonstrate the strongest correlation with the acentric factor and entropy.

The features of logistics network structures and prospects for their transformation in the BRICS countries

The ongoing global transformation processes have led to changes in logistics systems that are now being restructured at a new technological level. The formation of new inter-country groupings and the leap in the volume and quality of infrastructure have a key impact on logistics parameters; the increases in risk and uncertainty make it necessary to diversify supply capabilities and create reserve capacities in order to balance unforeseen situations. The paper first examines the formation of new logistics networks and diversification of existing ones using the evidence both from the BRICS countries, primarily Russia, India and China, and from candidates for accession to this organization; then, it analyses the prospects of their integration in the new geopolitical and technological environment. The study is based on the systematic approach, using comparative analysis and statistical methods; it aims to assess the prospects for cooperation between Eurasian members of BRICS in the field of logistics systems and to identify the national specifics of their financing, design and formation. The paper systematizes information on industry, maritime, air and land transport, paying special attention to the shipping of hydrocarbons and developing of latitudinal and meridional corridors on a new technological basis. It also gives an overview of the benefits gained from complex integrated initiatives and transformation of the logistics industry as a whole.

Gender Segregation in Civic Life: Women’s and Men’s Involvement in Voluntary Associations

The pervasive persistence of gender segregation has been documented in a myriad of social settings, implying that women and men primarily encounter peers of their own gender in daily life. While voluntary associations are often praised for their ability to bridge other social divides, previous research indicates substantive gender disparities in voluntary involvement. Yet we still know relatively little about the extent and origins of gender segregation in civic life. In this article, we study gender homophily in voluntary involvement and examine how structural features of friendship networks and traditional gender norms bring about gender segregation. Employing data from a German panel study (SC6-NEPS), we analyze cross-sectional patterns of gender segregation and run multinomial and binary logistic regressions to model joining and quitting transitions. Our results indicate substantive gender segregation across and within types of voluntary associations. The overall gender segregation is driven by homophilous transitions into associational contexts, not by selective quitting decisions. Gender-segregated friendship networks partially explain the tendency to join organizations dominated by one’s own gender. Traditional gender norms are associated with more homophilous joining transitions among men, but not among women. Overall, these findings imply that civic life perpetuates the structural significance of gender.