Complexity最新文献

Violence in Mexico presents a significant challenge to public safety and social stability and requires a comprehensive understanding of its temporal dynamics. This study uses a complex temporal networks approach to analyze patterns of violence across municipalities, focusing on crimes such as homicide, sexual assault, discrimination, and armed robbery. Using data spanning 6 years (2015–2021), this study builds temporal networks to find fundamental patterns and trends in violence. Then, using advanced network analysis techniques, including robustness assessment and influence quantification, key influential municipalities were identified within the violence network. The findings of this research can offer valuable information to carry out specific interventions and resource allocation strategies to improve public safety in Mexico.

Nonlinear science represents a foundational frontier in scientific inquiry that explores the shared characteristics inherent in nonlinear phenomena. This study focused on the perturbed Boussinesq (PB) equation incorporating dual perturbation terms. Soliton solutions were deduced by leveraging the traveling wave hypothesis. Furthermore, by employing the generalized Jacobi elliptic expansion function (JEEF) method and the improved tan (Λ/2) method, diverse nonlinear wave solutions, including kink, dark, periodic, bright, singular, periodic waves, bell-shaped solitons, solitary waves, shock waves, and kink-shaped soliton solutions, were acquired. The establishment of constraint relations is detailed to delineate the criteria for the existence of these wave solutions. Notably, these solutions are innovative and present novel contributions that have not yet been documented in the literature. In addition, 2D and 3D graphics were constructed to visually elucidate the physical behavior inherent to these newly acquired exact solutions.

In this work, the population dynamics of psychoactive substance abuse model governed by a system of nonlinear ordinary differential equations is studied through the efficacy and economic analyses. The study considers three time-variant intervention strategies, such as screening of addicted immigrants, advocacy against substance abuse and rehabilitation therapy for the control of substance abuse spread in the population. Most sensitive parameters to be targeted by the control interventions are examined through sensitivity analysis using the normalized forward sensitivity index. By employing optimal control theory, the existence of optimal control is qualitatively analysed and the control triple is characterized using Pontryagin’s maximum principle. To further assess the impact of the intervention strategies, three different policies combining any two of the intervention strategies, namely, Policy A (combination of screening of addicted immigrants and advocacy against substance abuse), Policy B (combination of screening of addicted immigrants and rehabilitation therapy) and Policy C (combination of advocacy against substance abuse and rehabilitation therapy), are considered. Efficacy and economic analyses are extensively conducted on each of the policies, and it is revealed that Policy B is the most efficient intervention, while Policy C is the most cost-effective intervention. As a result, the policy that averts the highest number of psychoactive substance abuse cases when limited resources are available is recommended.

Bridge inspection text records are very significant for the maintenance and upkeep of bridges, which can help engineers and maintenance personnel to understand the actual condition of bridges, detect and repair problems in time, and ensure the safe operation of bridges. Currently, more and more research focuses on how to extract potentially valuable bridge-related information from bridge inspection texts. In this study, we take the bridge inspection domain machine-reading comprehension corpus as the data support for model training and performance evaluation; oriented to the bridge inspection domain data text extraction machine-reading comprehension task, on the basis of word-granularity text input, we further explore two schemes of co-occurring linkage of cross-sentence entities in the context and co-occurring linkage of entities within the sentence through graph structure, and we learn and extract naming through graph-attentive neural networks-structured semantic information between entities and fused the obtained named entity embeddings with a hidden representation of the pretrained context. Tested on the bridge inspection domain dataset, the integrated model proposed in this research improves the EM optimum by 1.4% and the mean by 2.2% and the F1 optimum by 2.2% and the mean by 1.6% on the BIQA test, compared with the better-performing baseline model RoBERTa_wwm_ext.

The widespread use of the internet has led to the rapid expansion of social networks, making it easier for individuals to share content online. However, this has also increased the prevalence of cyber violence, necessitating the development of automated detection methods. Deep learning-based algorithms have proven effective in identifying violent content, yet existing models often struggle with understanding contextual nuances and implicit forms of cyber violence. To address this limitation, we propose a novel deep multi-input recurrent neural network architecture that incorporates neighborhood-based contextual information during training. The Jaccard similarity metric is employed to construct neighborhoods of input texts, allowing the model to leverage surrounding context for improved feature extraction. The proposed model combines Bi-LSTM and GRU networks to capture both sequential dependencies and contextual relationships effectively. The proposed model was evaluated on a real-world cyber violence dataset, achieving an accuracy of 94.29%, recall of 81%, precision of 72%, and an F1-score of 76.23% when incorporating neighborhood-based learning. Without contextual information, the model attained an accuracy of 89.15%, recall of 72.00%, precision of 71.5%, and an F1-score of 71.74%. These results demonstrate that neighborhood-based learning contributed to an average improvement of 5.14% in accuracy and 4.49% in F1-score, underscoring the importance of contextual awareness in cyber violence detection. These results highlight the significance of contextual awareness in deep learning-based text classification and underscore the potential of our approach for real-world​ applications.

This study explores the dynamics of a discrete-time predator–prey system, incorporating a Holling type III functional response and prey refuge, through the piecewise constant argument method. This method keeps things consistent and prevents negative population values, which is often a drawback of older techniques. We take a closer look at fixed points, exploring their existence and stability. We identify the conditions that lead to period-doubling and Neimark–Sacker bifurcations, and we back up our findings with numerical simulations. Our findings emphasize how important the predation coefficient is for keeping ecological balance and show that, in this model setup, the refuge for prey has a minimal effect on the stability of the system. These insights help us better understand the relationships between predators and their prey, providing valuable guidance for conserving biodiversity and managing ecosystems.

This paper proposes a deterministic nonlinear epidemic model, $$, incorporating media coverage for the analysis of infectious disease transmission dynamics, considering quarantine and isolation control strategies in a community with pre-existing immunity, and taking into account both asymptomatic and symptomatic infections. The model examines two equilibria: the disease-free equilibrium and a unique endemic equilibrium, investigating their existence and local stability concerning the effective reproduction number, R_e. In addition, global stability analysis of these equilibria is conducted. Media coverage is found to have no effect on R_e but plays a crucial role in mitigating the disease burden by reducing the number of infectious individuals at the endemic steady state and decreasing the infection peak. A new method is introduced to estimate the coefficients of media coverage. Based on the findings of central manifold theory, it is observed that when R_e exceeds unity, a transcritical bifurcation takes place in the system, leading to the asymptotic stability of the endemic equilibrium. Moreover, it is noted that the effectiveness of quarantine and isolation at the population level depends on the transmission rate among isolated and quarantined individuals, while higher pre-existing immunity in the population reduces the infection peak and hastens its occurrence. Theoretical results are validated through simulations, and sensitivity analysis is conducted for the effectiveness of R_e with respect to model parameters.

This study introduces a novel evolutionary game-based model to explore the dynamic interplay between venture capital (VC) institutions and their invested firms in adopting corporate social responsibility (CSR) practices, with a focus on the moderating role of government supervision. With CSR gaining prominence in VC-backed firm ecosystems, understanding these dynamics is critical. Verified through numerical simulations, the model examines system evolution under scenarios with and without government supervision. Key findings include the following: (1) VC-backed firms are more likely to adopt CSR practices when VCs prioritize long-term investment strategies, whereas short-term strategies centered on shareholder value hinder CSR efforts; (2) factors such as VC equity share, CSR costs, and benefits significantly influence CSR adoption; and (3) government incentives can promote CSR and accelerate convergence toward ideal outcomes, though with limited impact. This research advances the understanding of the VC-CSR relationship, offering theoretical insights and practical recommendations to foster responsible investment and business practices.

Enterprise sponsored virtual communities (ESVC) are important platforms for user knowledge sharing. User characteristics, as influential factors in these communities, impact the evolution of knowledge sharing processes. In this study, a knowledge sharing evolutionary game model that considers the influence of user characteristics was constructed. A payoff matrix was established, and stable points were identified for further analysis. Simulation analysis was conducted to explore the impact of changes in user characteristics on the knowledge sharing strategies of both parties. This study revealed the following findings: First, the knowledge reservoir of ESVC users affects their willingness to share knowledge, with ordinary users’ willingness initially increasing and then decreasing as their knowledge reservoirs grow. Second, the willingness of leading users to share knowledge initially decreases and then increases as their knowledge reservoirs expand. Third, improvements in the knowledge absorption and transformation capabilities of both leading and ordinary users promote knowledge sharing behaviour, although the extent of this impact differs between the two groups.

In recent years, there has been a marked surge in research focused on nonlinear oscillators. Among these, a particular emphasis has been placed on a class of oscillators distinguished by their concealed attractors, drawing considerable attention due to their unique characteristics. This paper delves into the exploration of an elegant oscillator belonging to this distinctive class. Despite comprising five terms and lacking equilibrium points, this oscillator displays remarkably intricate dynamics. The study covers various aspects such as chaos, hidden attractors, offset boosting, and notably, different strange attractors exhibited by this oscillator. Additionally, approaches involving synchronization for such oscillators are introduced. Apart from the presentation of the novel chaotic oscillator, the synchronization of a nominal and uncertain chaotic system is evinced by the sliding mode technique (super-twisting algorithm) in the first case, and a robust controller is synthesized, respectively. The appropriate Lyapunov functions are implemented in the two synchronization strategies leading to obtain suitable control strategies to achieve fast and accurate control laws. The respective simulations are performed along with the conclusions of this work.