Narrow Results

This paper presents a novel link prediction approach, termed Basic-Structural Similarity Link Prediction (BSSLP), designed to address the zero-similarity problem in complex networks. BSSLP integrates basic similarity, which establishes a nonzero baseline for all node pairs, with structural similarity that captures both local and intermediate topological features. This integration effectively mitigates challenges such as cold start and sparse network prediction. Through extensive experiments on nine real-world networks, BSSLP consistently outperforms seven benchmark methods, achieving an average AUC improvement of 5.17%. The method demonstrates robust performance across various network structures and maintains high prediction accuracy under different training set proportions. By providing nonzero similarity estimates for all potential edges, BSSLP significantly enhances the prediction of new connections and offers deeper insights into network dynamics.

Knowledge-sharing behavior based on cluster innovation network has become the primary way for enterprises to achieve sustainable innovation. The willingness of enterprises to share knowledge is a dynamic evolutionary game, supplemented by their periodic evaluations of changes in the network structure and strategic optimization based on game payoffs. An evolutionary game model is constructed to determine which network structure enables enterprises to attain the highest willingness to share knowledge. The results show the following: (1) The scale-free network exhibits the most significant influence on enhancing enterprises’ willingness to share knowledge, followed by small-world, random, and regular networks. In addition, the scale-free network demonstrates the slowest rate of evolutionary equilibrium. (2) The more apparent the agglomeration and community structure within a network, the greater the likelihood that enterprises will adopt knowledge-sharing strategy. (3) Compared to the network structure with smaller communities, larger communities can enhance enterprises’ willingness. However, larger communities cannot be substituted by an equivalent number of smaller communities, even if they contain the same number of enterprises. The network evolutionary mechanism will be clarified, contributing to the optimization of network structure and the promotion of knowledge-sharing collaboration among enterprises, which facilitates cluster innovation.

This paper adopts the visibility graph (VG) methodology to analyze the dynamic behavior of West Texas Intermediate (WTI), Brent and Shanghai (SC) crude oil futures during the COVID-19 pandemic and Russia–Ukraine conflict. Utilizing daily and high-frequency data, our study reveals a clear power-law decay in VG degree distributions and highlights pronounced clustering tendencies within crude oil futures VGs. We also uncover an inverse correlation between clustering coefficients and node degrees, further identifying that all VGs adhere not only to the small-world property but also exhibit intricate assortative mixing. Through the time-varying characteristics of VGs, we observe that WTI and Brent demonstrate aligned behaviors, while the SC market, with its unique trading mechanisms, deviates. Notably, the five-minute assortativity coefficient provides deep insights into the markets reactions to these global challenges, underscoring the distinct sensitivity of each market.

We show how the connection structure of a loopless communication network may be discovered using only ubiquitous echo requests or as a byproduct of normal two-way transport. The key factor is the correlation effect in waiting times of successively sent messages, which is caused by background traffic on the routers.

In this paper, we propose an efficient strategy to enhance the network transport efficiency by adding links to the existing networks. In our proposed strategy, we consider both the node betweenness centrality (BC) and the shortest path length (L) as two important factors. The overall traffic capacity of a network system can be evaluated by the critical packet generating rate R_c. Simulation results show that the proposed strategy can bring better traffic capacity and shorter average shortest path length than the low-degree-first (LDF) strategy and the low-betweenness-first (LBF) strategy. This work is helpful for designing and optimizing of realistic networks.

We analyze the impact of the network structure, the default probability and the loss given default (LGD) on the loss distribution of systemic defaults in the interbank market, where network structures analyzed include random networks, small-world networks and scale-free networks. We find that the network structure has little effect on the shape of the loss distribution, whereas the opposite is true to the default probability; the LGD changes the shape of the loss distribution significantly when default probabilities are high; the maximum of the possible loss is sensitive to the network structure and the LGD.

Car-sharing program, like Car2go, is an innovative urban transportation mode where the car-sharing company provides a car fleet to offer people with the short-term access of car traveling. As a new traveling service, car-sharing platforms have been struggling hard to trigger initial users and speed up their diffusion process. Unlike new product spreading via geographical proximity people, car-sharing users usually drive sharing cars to different destinations and influence people there, and potential user decision also depends on previous user activity at all their destinations. Car-sharing user connections are mainly affected by their traveling behaviors. The influence of user traveling network on new service/product spreading process has been rarely studied before. Here, we find that the infective rate between users with the same destination is critical to the minimum user base of car-sharing diffusion. Moreover, a city with central user network is more appropriate for car-sharing. It leads to a small critical infective rate for diffusion, and a large stable market size of car-sharing service. Our study can impact car-sharing market strategies ranging from market expansion in one city to optimal market selection among different cities.

The structural and dynamical properties in sodium silicate liquid were investigated by molecular dynamics method. To clarify the distribution of sodium atoms in model, characteristics of simplex have been investigated. The simulation results reveal that Na₂O⋅4SiO₂ (NS₄) liquid has a lot of simplexes with four sodium atoms inside but about half of simplexes do not have sodium. The spatial distribution of sodium is nonuniform, sodium tends to be in the nonbridging oxygen-simplexes and in larger-radius simplex. Moreover, the sodium density for nonbridging oxygen region is significantly higher than the one for Si-region. Namely, link-cluster function F${}_{lk}$(r, t) has been used to clarify dynamical heterogeneity in NS₄ liquid. The F${}_{lk}$(r, t) for sets of random, immobile and mobile network atoms is quite different, which indicates that the dynamics of network atoms is heterogeneous. The Si–O network has the structure with two separated domains (immobile and mobile domains). These types of domain are significantly different in local microstructure, mobility of atoms and chemical composition.

This study reported a simulation of structural transition and correlation between structural and dynamical heterogeneity (DH) for liquid Al₂O₃. Structural characteristics of liquid Al₂O₃ were clarified through the pair radial distribution functions, the distribution of ${\text{AlO}}_x$ and ${\text{OAl}}_y$ ($x=3$, 4, 5, 6; $y=1$, 2, 3) basic structural units, angle and bond length distribution and 3D visualization. Simulation results revealed that network structure of liquid Al₂O₃ is built mainly by AlO₃, AlO₄, AlO₅ and AlO₆ units that are linked to each other through common oxygen atoms. We found the existence of separate AlO₄-, AlO₅- and AlO₆-phases where the mobility of atoms can be determined. The atoms in AlO₄-phase are more mobile than the ones in AlO₅- and AlO₆-phases. The existence of separate phases is evidence of DH in liquid Al₂O₃. Moreover, the self-diffusion of Al and O atoms was also discussed via characteristics of separate AlO₄-, AlO₅- and AlO₆-phases.

The traffic dynamics of complex networks is closely related to network structure. By changing network structure, the traffic dynamics behavior can be optimized. Faced with the network congestion problem, we focus on the relationship between network traffic capacity and its structure. The multilayer networks are studied, which are composed of high-speed and low-speed layers. A link rewiring strategy is proposed to change the low-speed layer structure and improve the network traffic capacity. Compared with the random link rewiring strategy, the purposeful link rewiring strategy can improve network traffic capacity. A large number of simulations are carried out under the effective traffic-flow assignment strategy to prove the effectiveness of the link rewiring strategy. This strategy improves packet transmission efficiency of low-speed layer, and reduces the average length of effective path, which indicates that adjustment of low-speed layer structure can improve traffic capacity of multilayer networks.

Considering the heterogeneous structure of scale-free networks causing low traffic capacity of network, we propose to improve the network transport efficiency by rewiring a fraction of edges for the network. In this paper, six edge rewiring strategies are discussed and extensive simulations on Barabási–Albert (BA) scale-free networks confirm the effectiveness of these strategies. From another perspective, rewiring edges for scale-free networks directly reuse the removed edges under some edge-removal strategies [Z. Liu, M. B. Hu, R. Jiang, W. X. Wang and Q. S. Wu, Phys. Rev. E76 (2007) 037101; G. Q. Zhang, D. Wang and G. J. Li, Phys. Rev. E76 (2007) 017101], and can significantly enhance the traffic capacity of the network at the expense of increasing a little average path length. After the edge rewiring process, the network structure becomes significantly homogeneous. This work is helpful for network design and network performance optimization.

We perform a simulation of the structural phase-transition pathway under compression and dynamic properties in liquid germania (GeO₂). The structure of liquid GeO₂ is clarified through the pair radial distribution function (PRDF), distribution of GeO${}_x$$(x=4,5,6)$ units, bond angle and length distribution, and three-dimensional (3D) visualization. The result shows that the structure of liquid GeO₂ is built by GeO₄, GeO₅ and GeO${}_6$units, which are linked to each other via common oxygen atoms. The GeO${}_x$ units lead to form into the separate GeO₄-, GeO₅- and GeO₆-phases. The existence of separate phases is evidence of dynamical heterogeneity (DH) in liquid GeO₂. The atoms in GeO₅-phase are more mobile compared to other ones. The variation of the self-diffusions of Ge and O atoms under pressure is examined via the characteristics of separate GeO₄-, GeO₅- and GeO₆-phases. We found that under compression, there is diffusion anomaly in liquid GeO₂. This is suggested to be related to the very high mobility of Ge and O atoms in the GeO₅-phase compared to GeO₄- and GeO₆-phase.

Using the fact that connections between vertices of a network often represent directed and weighted flows, we apply hydraulic principles to develop novel insights into network structure and growth. We develop a network model based on Bernoulli's principle and use it to analyze changes in network properties. Simulation results show that velocity of flow, resistance, fitness and existing connections in a system determine network connections of a vertex as well as overall network structure. We demonstrate how network structure is affected by changes in velocity and resistance, and how one vertex can monopolize connections within a network. Using Bernoulli's principle, we are able to independently reproduce key results in the network literature.

In this paper, we investigate how contagion risk is affected by bank activities in four types of interbank network structures, that is, random, small-world, scale-free and tiered networks. We vary the key parameters that define bank activities in the interbank market — including the size of interbank exposures, the size of liquid assets, the heterogeneity of the size of credit lending and the heterogeneity of banks — and analyze the impact of these parameters on contagion risk. First, we find that the size of interbank exposures is the main factor in determining the effect of contagion risk, that increases in the size of interbank exposures may lead to an increase in the threat of contagion risk, that after the size of interbank exposures rises beyond a threshold, the effect of contagion risk in small-world networks is the most significant, followed by that in tiered, random and scale-free networks, respectively. Second, increases in the size of liquid assets can decrease the effect of contagion risk. Third, the impact of the heterogeneity of the size of credit lending on contagion risk varies with interbank network structures. Finally, the effect of contagion risk among heterogeneous banks is stronger than that among homogeneous banks, and there is a positive relationship between the effect of contagion risk and the heterogeneity of banks.

This paper seeks to improve our understanding of how network structure affects innovation outcomes in strategic innovation networks. The theoretical argument is illustrated by the case of a Swedish strategic innovation network. Asocial network analysis of the relationships between 58 network members was conducted. Roughly half of the network actors were involved in producing innovations; they had significantly larger, higher-density networks and occupied more central positions in their networks than did those not participating in the innovation or scientific work.

Cu–$X$ CNTs composites ($X=0$, 1, 2, 3 vol.%) were successfully prepared using a combination of pre-treatment, powder metallurgy and multi-directional forging processes, which provides a solution for the industrial manufacture of the composites with network CNTs structures. During the multi-directional forging process, the CNTs in the composites were distributed in a network under the synergy of metal flow and copper particle squeeze. Compared with other structure modes, the network CNTs can effectively carry and transfer loads resulting in the promotion of mechanical properties (such as, the tensile strength approximately 1.5 times higher than those of composites with the same volume fraction without network structure). The composite with 2 vol.% CNTs had the highest elongation in this experiment (41%), which is about 5 times higher than the composites with other CNTs distribution patterns. At a low CNTs content level (1vol.%), a complete load transfer network cannot be formed, resulting in a relatively insufficient mechanical properties of the composites. As the content level is exceeded (3vol.%), it caused significant agglomeration of the CNTs, which lead to fracture in the agglomerated CNTs and elongation degradation of the composites.

Cu-$X$ CNTs composites ($X=0$, 1, 2vol.%) were successfully prepared through a combination of pre-treatment, powder metallurgy and multi-directional forging (MDF) processes. It provided a solution for the industrial production of structural composites with reticulated CNTs. The effect of CNTs contents on wear and electrical performance of the composites was investigated. The addition of CNTs results in a significant increase in the wear resistance of the composites. During the wear process, reticulated CNTs will carry the load barrier to high temperature softening and plastic deformation of the Cu-matrix, which protects the matrix eventually. In general, the addition of CNTs will reduce the conductivity of the composites. Under MDF process, the CNTs were organically linked together and formed a network structure as the volume fraction increased from 1% to 2%, resulting in a significant promotion of conductivity by providing paths for electron transfer. The contact resistance increased slightly when 2vol.% CNTs were added, with an average value of 5.87m$\mathrm{\Omega }$. The contact resistance volatility of the Cu-2vol.% CNTs was the most stable (variance of 1.485). Reticulated CNTs can effectively attenuate the erosive effect of molten pools and welding bridges on the anode material, and are ideal reinforcers for Cu-based electrical contact materials.

The cooperation among cities to promote scientific and technological (sci-tech) innovation is of practical significance to regional coordinated development. Therefore, it is necessary to explore the features of sci-tech cooperation and innovation network in China’s urban system. Based on the recognition that high-level scientific papers are important achievements of sci-tech innovation, the author investigates the status quo of the cooperation on high-level scientific papers among cities, establishes a matrix on sci-tech innovation network in China’s urban system and analyzes its structural features and evolutionary trend during 2000–2010. The results show that: urban sci-tech innovation network is developing rapidly, but the development level remains low; network dominant city pairs are distributed across regions; network connection keys, concentrating in the eastern cities, present clear regional differences in distribution; the structure of network dominant cities is relatively stable, but the first network node city (Beijing) is underdeveloped; cohesive subgroups start to develop, but neither the national nor the regional high-level cohesive subgroups are fully developed; the largest subgroup consists of only 25 cities; and none of the three major urban agglomerations (Beijing–Tianjin–Hebei, Yangtze River Delta, and Pearl River Delta urban agglomerations) forms a complete cohesive subgroup.

In this study, we construct a green innovation network by using the gravity model and propose a panel model to analyze the mechanism by which the green innovation network structure of the urban agglomeration affects eco-efficiency and take the Beijing-Tianjin-Hebei urban agglomeration as an example. Results show that the urban agglomeration of Beijing, Tianjin, and Hebei is characterized by radial development, with “Beijing-Tianjin” serving as the core. The central position of city nodes in the innovation network of the Beijing-Tianjin-Hebei urban agglomeration significantly impacts its eco-efficiency, whereas other characteristics do not.

According to the trade-off between information diffusion and diversity in an efficient network, we extend Lazer's simulation model on parallel problem solving by adding partner selection strategy: structurally equivalent imitation. In this way we can examine how the interaction of network structure with agent behavior affects the knowledge process and finally influence group performance. Our simulation experiment suggests that when agents adopt structure equivalence imitation the whole organization implicitly would be divided into independent sub-groups which converge on the different performance level and lead the whole group to a lower performance level.