Network Theory Research Articles

Complex-Network based model for SMS spam filtering

With the advancement of technology and the widespread use of mobile phones and wireless communication, SMS has become the most popular texting method due to its high response rate, affordability, and no internet connection requirement. A survey found that 3.5 billion users, or 80% of active users worldwide, use SMS for communication. SMS, however, has also attracted spammers, resulting in an explosion in spam messages, especially in Asia. Users are annoyed, lose money, and waste their time by receiving spam messages intended to serve various purposes, such as advertising, adult content, smishing, and fraud. Spam messages are a problem for users and providers, which calls for a mechanism to identify and filter them out. With supervised machine learning techniques, we propose a novel approach to classify spam and ham messages based on complex network theory. The proposed approach integrates complex network based features with statistical TF-IDF and grammatical rules features. Also, an under-sampling method has been employed in order to cope with the imbalanced data issue. We evaluated the performance of several supervised learners in terms of accuracy, precision, recall, F1-score, and AUC. In our experiments, Random Forest successfully classified spam messages more accurate than statistical methods that only extracted TF-IDF features.

How large-scale sport events redefine rurality? CunBA and the (re)production of rurality in China

The rural revitalization experience in China is being transformed by the village basketball association (CunBA), an event embedded in rural everyday practices or strategically designed to shape the perception of rural space. As this event integrates into daily rural life, it has the potential to alter both rural space and rurality itself. Using the theoretical framwork of performing rurality and employing Actor Network Theory (ANT) as an analytical tool, this paper examines the process and effects of rurality transformation as various actors, including the state and digital platform users, become part of this network. Through the analysis of semi-structured interviews and netnography data, the study reveals that the reconstruction of rurality is influenced by the interactions within a heterogeneous actor network. It finds that rurality is enacted within a dynamically equilibrated network and that the meaning and outcomes of rurality are redefined through the interactions between human and non-human actors and the evolving network. By focusing on rural China, this paper aims to contextualize the concept of ‘performing rurality’ and further explore its theoretical issues in a digital context. The findings indicate that rural revitalization in China increasingly depends on digital platforms to perform rurality. This phenomenon must be understood within the context of ‘post-rurality’ and the broader socio-cultural processes behind it in rural China.

Quantitative risk assessment of road exposed to landslide: A novel framework combining numerical modeling and complex network theory

The quantitative analysis of the landslide risk posed to road networks is a challenging task owing to the uncertainty involved both in the potential landslide hazard and the road value. To address this challenge, this paper proposes a novel framework to assess the road risk in quantitative terms. The landslide hazard is assessed using the depth-integrated method with consideration of the landslide size probability and initial conditions. The potential direct losses associated with road disruption are determined by exposure analysis in a geographical information system. The indirect losses of the road caused by the landslide were analyzed through complex network theory with consideration to regional socioeconomic development and the time required for road restoration. The proposed framework was used to assess the road risk posed by the Chunchangba landslide, Xiaojin County, China. The results show that the volume size probability of landslides in the Xiaojin area could be assessed using the function PVL=1/1+3.583×10−5VL2×0.391. The depth-integrated method based hazards assessment results show the maximum impact area of the Chunchangba landslide reached 0.383 km2, and the landslide has a high probability of damaging the lower road and forming a barrier lake. The losses associated with road disruption caused by the landslide were estimated. The maximum direct losses reached 293,100 USD, while indirect losses reached 423,800 USD, which has the same importance as direct losses. The risk curve reveals that the maximum probability of the road risk associated with the Chunchangba landslide is 0.0175 %, 0.0189 %, 0.0190 %, and 0.0191 % for the time interval of 5, 10, 20, and 50 years, respectively, with losses of 0.177 million USD/year. The regional disaster mitigation strategy is analyzed based on quantitative risk analysis. The results show that a new 2.7 km road on the mountain opposite the Chunchangba landslide can reduce indirect losses by approximately 300 times. The findings of this study contribute to sustainable development and landslide risk management in mountainous areas.

A reaction network model of microscale liquid-liquid phase separation reveals effects of spatial dimension.

Proteins can form droplets via liquid-liquid phase separation (LLPS) in cells. Recent experiments demonstrate that LLPS is qualitatively different on two-dimensional (2D) surfaces compared to three-dimensional (3D) solutions. In this paper, we use mathematical modeling to investigate the causes of the discrepancies between LLPS in 2D and 3D. We model the number of proteins and droplets inducing LLPS by continuous-time Markov chains and use chemical reaction network theory to analyze the model. To reflect the influence of space dimension, droplet formation and dissociation rates are determined using the first hitting times of diffusing proteins. We first show that our stochastic model reproduces the appropriate phase diagram and is consistent with the relevant thermodynamic constraints. After further analyzing the model, we find that it predicts that the space dimension induces qualitatively different features of LLPS, which are consistent with recent experiments. While it has been claimed that the differences between 2D and 3D LLPS stem mainly from different diffusion coefficients, our analysis is independent of the diffusion coefficients of the proteins since we use the stationary model behavior. Our results thus give new hypotheses about how space dimension affects LLPS.

Holography of transmission lines: Insights of continuous MERA and AdS/CFT

This study examines the holographic representation of the quantum theory of transmission lines, which play a crucial role in quantum computing and quantum information. By utilizing Yurke and Denker’s quantum circuit network theory within the framework of continuous Multi-scale Entanglement Renormalization Ansatz (cMERA) in AdS space, we analyze the quantization and interactions of transmission lines. The metric is revealed to be described by the inductance of the quantum circuit, which is AdS space in its 0-limit. These results provide new insights into handling and controlling complex phenomena in quantum circuits, potentially advancing the understanding of quantum computing and quantum communication.

A Survey on Statistical Theory of Deep Learning: Approximation, Training Dynamics, and Generative Models

In this article, we review the literature on statistical theories of neural networks from three perspectives: approximation, training dynamics, and generative models. In the first part, results on excess risks for neural networks are reviewed in the nonparametric framework of regression. These results rely on explicit constructions of neural networks, leading to fast convergence rates of excess risks. Nonetheless, their underlying analysis only applies to the global minimizer in the highly nonconvex landscape of deep neural networks. This motivates us to review the training dynamics of neural networks in the second part. Specifically, we review articles that attempt to answer the question of how a neural network trained via gradient-based methods finds a solution that can generalize well on unseen data. In particular, two well-known paradigms are reviewed: the neural tangent kernel and mean-field paradigms. Last, we review the most recent theoretical advancements in generative models, including generative adversarial networks, diffusion models, and in-context learning in large language models from two of the same perspectives, approximation and training dynamics.

Brand, corporate and celebrity credibility: a reverse effect

PurposeCelebrity endorsers are usually considered to bring positive effects to associated nodes, such as brands and corporations. However, limited evidence suggests that brands and corporations are equally responsible for affecting celebrities and their credibility. Drawing on associative network theory, this study explores the effects of brand credibility and corporate credibility on celebrity credibility, both directly and through the mediating and moderating effects of advertising credibility. The research addresses three main issues: (1) whether brand credibility, corporate credibility and advertising credibility have significant effects on celebrity credibility; (2) whether advertising credibility has a significant mediating effect on the effects of brand credibility and corporate credibility on celebrity credibility and (3) whether advertising credibility has a significant moderating effect on the effects of brand credibility and corporate credibility on celebrity credibility.Design/methodology/approachThe study used a quantitative approach involving structural equation modelling. Data were collected from 675 participants from London and focussed on four leading international brands, corporations and celebrity endorsers.FindingsThe findings show that brand credibility and advertising credibility have positive direct effects on celebrity credibility; and that advertising credibility mediates the effects of both credibility constructs on celebrity credibility. Furthermore, moderating effects of advertising credibility are also found.Practical implicationsThis study will help managers to understand the reverse effects, i.e. the effects of brand credibility and corporate credibility on celebrity credibility. They will be able to understand that a credible brand and corporation like a credible celebrity can also bring significant effects on the associated elements. This will help them to recruit celebrity endorsers who have historically earned their credibility from previous endorsements of credible brands and corporations. Further, these findings will help managers to understand that credibility of the brand and corporation can also affect the credibility of the associated advertising, resulting in having a significant effect on the credibility of the celebrity. This on the consumers’ side will enhance their preferences, attitudes and behaviours, while for the corporation, it will enhance their economic and commercial performance.Originality/valueThis is the first study in the literature, where a conceptual model based on the reverse effects of both credibility constructs on celebrity credibility is examined, directly and based on the moderating and mediating effects of advertising credibility. Hence, the contributions to the literature are threefold: first, the study examines the reverse effect of celebrity endorsement, whereby the credibility of a brand or corporation is transferred to a celebrity endorser; second, the study examines the mediating and moderating effects of advertising credibility on this reverse effect and finally, associative network theory is used to examine the importance of the model.

Global Stock Markets Volatility Correlation Structure and Implication of Portfolio Based on Complex Network Theory

Global Stock Markets Volatility Correlation Structure and Implication of Portfolio Based on Complex Network Theory

Liquid Hopfield model: Retrieval and localization in multicomponent liquid mixtures

Biological mixtures, such as the cellular cytoplasm, are composed of a large number of different components. From this heterogeneity, ordered mesoscopic structures emerge, such as liquid phases with controlled composition. The competition of these structures for the same components raises several questions: what types of interactions allow the retrieval of multiple ordered mesoscopic structures, and what are the physical limitations for the retrieval of said structures. In this work, we develop an analytically tractable model for multicomponent liquids capable of retrieving states with target compositions. We name this model the liquid Hopfield model in reference to corresponding work in the theory of associative neural networks. In this model, we show that nonlinear repulsive interactions are a general requirement for retrieval of target structures. We demonstrate that this is because liquid mixtures at low temperatures tend to transition to phases with few components, a phenomenon that we term localization. Taken together, our results reveal a trade-off between retrieval and localization phenomena in liquid mixtures, and pave the way for other connections between the phenomenologies of neural computation and liquid mixtures.

Misery Needs Company: Contextualizing the Geographic and Temporal Link between Unemployment and Suicide

Despite long-standing evidence linking higher unemployment rates to increased suicide rates, a puzzling trend emerged in the United States after the Great Recession: suicide rates continued to rise even as unemployment rates declined. Drawing on theories of social networks and reference groups, we advance the concept of “sameness”—in this case, the extent to which an individual’s employment status aligns with the fate of others in one’s community—to clarify how unemployment rates influence suicide. Constructing a multilevel dataset of U.S. suicide deaths from 2005 to 2017, we find that while unemployed individuals face a higher risk of suicide compared to the employed, this gap diminishes in communities with high local unemployment rates. Moreover, the “sameness” effect extends beyond geographic contexts to temporal ones, as national unemployment spikes reduce suicide risk among the unemployed and diminish the importance of local sameness. Together, these findings suggest a mechanism of “situational awareness,” whereby local and national economic contexts shape the meaning of unemployment, shifting its interpretation from personal failure to system failure and reducing its stigma. Our article offers a novel framework for examining the effects of cross-level interactions in suicide research, highlighting the crucial role of culture as deeply intertwined with social network mechanisms in shaping contextual influence.

Policy Framing, Social Networks, and Teacher Perspectives in a Technology Rollout

Drawing on theories of framing and social networks, we examine how educators define policy goals in a technology rollout, and how this framing process shapes teacher attitudes for technology. We analyzed social network data and 37 interviews with district leaders, digital learning coaches, and school staff from six elementary schools in a mid-sized district in California. These data suggest that school leaders and teachers re-framed district reform messages in different ways depending on the density of dialogue in their networks and who had social influence. These framings supported teacher motivation for using technology in ways that aligned to or built upon the district's instructional goals, or prompted teachers to resist change and/or make compliance-oriented and superficial shifts in their practice. We argue that district leaders should attend to school context when implementing technology rollouts and proactively engage school actors to define the goals of technology. We caution that technology rollouts demand sudden and intensive shifts in teacher work roles and can foster resistance to change without adequate time and support.

Моделювання процесу ранжування у нейромережному класифікаторі об’єктів

As part of expert systems for various purposes, one of the basic ones is the decision support subsystem, which, in turn, requires the need for a procedure for classifying ob-jects. This is especially evident in intelligent medical diagnostic systems, which widely use artificial intelligence methods and tools. In this context, an approach involving mod-ern neurotechnology methods has proven to be effective at a high level. This paper con-siders a variant of the structural organization of a neural network classifier of objects as an improved model of the Hamming neural network. The peculiarity of this variant of the classifier is the expansion of its functionality by forming the ranks of the classified ob-ject in all defined classes. In the case of medical diagnosis, this means ranking all possi-ble diagnoses of a disease, i.e. determining not only the most likely diagnosis, but also the closest in rank to it. In fact, this will allow us to clarify the diagnosis, and thus im-prove the results of medical diagnosis. Accordingly, we simulated the classification pro-cess with the ranking of results, which corresponds to the classification with the realiza-tion of competition between the neurons of the competitive layer using negative-reverse (lateral) connections. This approach is basic in the theory of neural networks for deter-mining the winning neuron according to the WTA (Winner Takes All). Simulation model-ing of the classification variant was performed using specific biomedical data (eight symptoms) for the diagnosis of appendicitis (four diagnoses). The results of modeling the processes of neural network classification of objects with the formation of appropri-ate ranks are presented in the form of a table. They confirmed the correctness of the functioning algorithm for the considered classification model.

Research on the Optimization and Development of Low Carbon Energy Economy from the Perspective of Water Conservancy Engineering

In order to explore the optimal development path of low-carbon energy economy in universities, this paper proposes a new method based on complex network theory to describe the global evolution and transformation characteristics of multivariate time series correlation patterns. In order to more intuitively see the impact of product extraction inputs as endogenous factors on economic growth, this article explores the relationship between economic growth and product extraction inputs by constructing a Hamiltonian function and using optimization methods. When considering whether the extraction input share of products with different extraction ratios is different, this article found the relationship between the total extraction input of products with different extraction ratios and the input share of green and low-carbon products, and analyzed consumers' preferences; Extractive investment of products under budget constraints and policy constraints. In addition, this article uses the dynamic equation of household utility function and environmental quality to prove that the extractive input of products not only promotes carbon reduction, but also has a positive effect on promoting economic growth. The simulation results show that the low-carbon economic development model based on energy efficiency proposed in this paper can effectively improve the development path of low-carbon energy economy.

Adaptive neural control for a class of random nonlinear Markov jump multi-agent systems with full state constraints

This paper proposes a consensus control protocol based on the adaptive backstepping technique for a class of random Markov jump multi-agent systems (MASs) with full state constraints. Each agent is described by the high-order random nonlinear uncertain system driven by random differential equations, where the random noise is the second-order stationary stochastic process. First, a distributed tracking controller is designed for Markov jump MASs, effectively handling the interaction and coupling terms between agents. Second, an appropriate tan-type barrier Lyapunov function is selected to keep the agents’ states from the violation of constraint boundaries, and the unknown nonlinear terms with Markov jump parameters are approximated by neural networks (NNs) theory. Moreover, to address the differential explosion problem, the extended state observer (ESO) is first introduced instead of employing NNs approximation or command filtering techniques. Finally, the exponentially noise-to-state stability in mean square is proved rigorously through the Lyapunov method, which guarantees all signals of the closed-loop system are bounded in probability and the tracking error converges to a small neighborhood around zero. Simulations are given to demonstrate the feasibility of theoretical results.

Utilizing social network analysis methodology to assess innovative leadership in it organization

Social networking technology is a new and promising view of IT organization management, allowing for a deeper understanding of the organization’s internal relationships, structure, and social dynamics. This article examines modern approaches and prospects for developing social networking technologies in managing an IT organization based on innovative leadership. The article’s primary purpose is to study modern approaches and prospects for developing social networking technologies as applied tools in IT organization management based on innovative leadership. The article is a theoretical research work that defines the prospects for further research in organizational management. It is expected that the study’s results will reveal the importance of social networks in the context of innovative leadership and justify the effective use of social networking technologies to improve the performance of an IT organization. In addition, identifying the main components that influence the effectiveness of innovation leadership will enable IT organization leaders to effectively implement innovation leadership strategies and contribute to the successful integration of innovations in the organizational environment. The study of this topic is of practical importance. It can be used to develop human resource management strategies and improve the mechanisms for implementing innovations in the organizational environment. Moreover, the research will offer new insights into how social network structures can be optimized to foster a culture of continuous innovation. Given this, the article’s author seeks to consider and study the effectiveness of innovative leadership through the prism of social network theory since she assumes that the structure of the organizational social network can influence its effectiveness.

Research on optimizing the propagation path search of individualized products with multi-source design changes

Abstract The individualization and diversification of customer requirements are gradually affecting the product design and manufacturing process, and the production mode of enterprises is evolving towards customer-oriented mass individualized customization. The change in customer requirements is inevitable in the process of individualized product design, which is the key factor leading to the change in product design. This paper establishes the standard of individualized product parts and components association relationship through correlation analysis, and constructs the individualized product parts and components association network model based on the complex network theory; on this basis, comprehensively considering the influence of parts nodes and association relationship in the complex network, it calculates the intensity of change propagation through the node importance degree and the probability of change propagation, and carries out the search for multi-source design change propagation paths through the improved ant colony algorithm, and evaluates the influence of the change of customer requirement by taking the change rate of the network as the evaluation index, to help enterprises to decide the request.

Assessing spatial structure in marine populations using network theory: A case study of Atlantic sea scallop (Placopecten magellanicus) connectivity.

Knowledge of the geographic distribution and connectivity of marine populations is essential for ecological understanding and informing management. Previous works have assessed spatial structure by quantifying exchange using Lagrangian particle-tracking simulations, but their scope of analysis is limited by their use of predefined subpopulations. To instead delineate subpopulations emerging naturally from marine population connectivity, we interpret this connectivity as a network, enabling the use of powerful analytic tools from the field of network theory. The modelling approach presented here uses particle-tracking to construct a transport network, and then applies the community detection algorithm Infomap to identify subpopulations that exhibit high internal connectivity and sparse connectivity with other subpopulations. An established quality metric, the coherence ratio, and a new metric we introduce indicating self-recruitment to subpopulations, dubbed the fortress ratio, are used to interpret community-level exchange. We use the Atlantic sea scallop (Placopecten magellanicus) in the northwest Atlantic as a case study. Results suggest that genetic lineages of P. magellanicus demonstrate spatial substructure that depends on horizontal transport, vertical motility, and suitable habitat. Our results support connectivity previously characterized on Georges Bank and Mid-Atlantic Bight. The Gulf of St. Lawrence genetic lineage is found to consist of five subpopulations that are classified as being a sink, source, permeable, or impermeable using quality metrics. This approach may be applied to other planktonic dispersers and prove useful to management.

Design of a CPW-based SSPP band-pass filter with reflectionless notch

Abstract This work proposes a band-pass filter (BPF) with a reflectionless notch based on spoof surface plasmon polaritons (SSPPs), utilizing interdigital coupling structures and novel transmission line unit cells. This filter efficiently transmits signals within the 0.67 GHz −4.06 GHz frequency range. By analyzing the equivalent LC circuit of the novel transmission line unit cell, its dispersion relation is derived using microwave network theory, with a cutoff frequency of 4.11 GHz. By comparing its dispersion relation with that of the conventional transmission line unit cell, the miniaturization capability of the proposed unit cell can be verified. In the BPF, loading a zigzag groove onto the central transmission line can be equivalently represented as an interdigital coupling structure, generating a stopband in the low-frequency range near 0 GHz. By deriving and analyzing its S-parameters, it is shown that the bandwidth of the low-frequency stopband can be flexibly adjusted by modifying the geometric dimensions of the zigzag groove. Additionally, loading another set of interdigital coupling structures onto the transmission line generates a notch at its resonant frequency of 3.4 GHz. It is noteworthy that in this configuration, the interdigital coupling structures along with the central transmission line can be represented equivalently as a set of CPW antennas. At the resonant frequency, the atructure radiates signals into free space, forming a reflectionless notch. Based on the simulations, a physical filter was fabricated and tested, showing excellent agreement between simulations and measurements.

Toward an improved causal test of network effects: Does alliance network position enhance firm innovation?

AbstractResearch SummaryAssessing whether network position causes firm outcomes is difficult because networks result from firms' strategic choices. To address this issue, we develop a research design that separates self‐driven network change from other‐driven change. Because other‐driven change suppresses the focal firm's agency, outcomes resulting from it can be interpreted as causal—under verifiable assumptions. We apply this method to reassess whether alliance network position affects firm innovation. Using data from the biotechnology industry, we find that structural holes (but not closure) increase firm innovation when self‐driven network change is involved but not when other‐driven change is involved. This raises two possibilities: network effects spuriously capture unobservable firm characteristics or network theories must be updated to account for agency.Managerial SummaryWe examine a crucial question for managers making decisions about alliance networks: Does a firm's position in its alliance network drive innovation, or is it a reflection of the firm's inherent qualities? Our research explores this question by examining network changes outside a firm's control. We find that bridging disconnected partners (structural holes) enhances innovation, but only when firms actively create these positions and not when these positions result from others' actions. Our findings suggest that network position alone does not drive innovation. Acknowledging that network position, per se, is not advantageous provides nuanced insight to better guide managerial decisions.

A statistical mechanics investigation of unfolded protein response across organisms

Living systems rely on coordinated molecular interactions, especially those related to gene expression and protein activity. The Unfolded Protein Response is a crucial mechanism in eukaryotic cells, activated when unfolded proteins exceed a critical threshold. It maintains cell homeostasis by enhancing protein folding, initiating quality control, and activating degradation pathways when damage is irreversible. This response functions as a dynamic signaling network, with proteins as nodes and their interactions as edges. We analyze these protein-protein networks across different organisms to understand their intricate intra-cellular interactions and behaviors. In this work, analyzing twelve organisms, we assess how fundamental measures in network theory can individuate seed proteins and specific pathways across organisms. We employ network robustness to evaluate and compare the strength of the investigated protein-protein interaction networks, and the structural controllability of complex networks to find and compare the sets of driver nodes necessary to control the overall networks. We find that network measures are related to phylogenetics, and advanced network methods can identify main pathways of significance in the complete Unfolded Protein Response mechanism.