Hasil untuk "Information technology"

Babu S., Rama Narayanan V.

Financial fraudsters use Gen AI, digital channels, global networks, and synthetic identities making it complex to identify the fraudulent activities. Traditional rule-based systems relying on traditional methods do not identify frauds which use multi-step transaction routing with multiple institutions and across borders. Graph database using Labelled Property Graphs, represents customers, accounts, and transactions as interconnected nodes and edges. By ingesting live transaction data, they apply pattern-matching and community-detection to expose suspicious subgraphs. Money-laundering rings or collusive clusters—and let investigators trace multi-hop links to “hub” accounts with clear visual audit trails. Machine learning models trained on vast historical datasets, using supervised classifiers (e.g., gradient boosting) and unsupervised anomaly detectors. Features like transaction amounts, geolocation consistency, device fingerprints, and temporal sequences feed these models, while recurrent architectures capture evolving fraud tactics. Yet they often suffer from concept drift, require extensive labelled data, underperform on imbalanced cases, and behave as opaque black boxes, generating false positives and hampering trust. A hybrid framework combines relational graph insights with statistical scoring, boosting detection accuracy, reducing false alarms, and enhancing investigators’ confidence in fraud detection and prevention.

Byung-Ju Sohn, Jihoon Ryu, Sang-Wook Yeh et al.

Abstract Understanding the relationship between tropical heavy rainfall and large-scale circulation provides valuable insights for improving the climate models. Here we use Gaussian Mixture Model to identify two distinct types of heavy rainfall over the tropical Pacific, “strong deep convection” and “moderately strong deep convection,” using satellite-borne precipitation radar measurements. They differ in two typical climatological deep convection-related rainfall modes between the western and eastern Pacific regions. The occurrence frequency of moderately strong deep convection is significantly different between the western and eastern Pacific, potentially linked to the Walker circulation. The enhanced Walker circulation appears to weaken the local Hadley circulation, thereby reducing strong deep convective activity in the eastern Pacific. This increases moderately heavy rainfall and decreases diabatic heating, which can affect global climate. We propose incorporating the close link between large-scale Walker circulation and mesoscale heavy convective rainfall into the current climate models.

Elsadig Musa Ahmed, Khalid Eltayeb Elfaki, Eimad Abusham

This study introduces a novel Hybrid Deep Ensemble (HDE) model, which can maximize the accuracy of the prediction by combining the benefits of multiple learning architectures to examine the predictive performance of Machine Learning (ML) Models. The proposed model consists of three models: Multiple Linear Regression (MLR), Random Forest Regression (RF), and Gradient Boosting Regression (GBR) as the base learners. Meta-learners combine these models' outputs to make final predictions using a regression model. The HDE model was applied to forecast Gross Domestic Product (GDP) and Carbon Dioxide (CO2) emissions by examining the influence of labor, capital, energy efficiency, and renewable energy in selected Asian Economies. The proposed HDE model performance is evaluated against two ensemble benchmark models: RF, which is based on bagging, and GBR, which is based on boosting; and MLR, which is a non-ensemble baseline algorithm. Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) measures were employed to evaluate the accuracy of the models using a dataset of economic and environmental indicators. With the lowest MAE and RMSE values for both GDP and CO2 emissions estimates, the results show that HDE is always lower than the MAE and RMSE values of both GDP and CO2, revealing the better suitability to predict complex and nonlinear patterns. This study highlights the importance of selecting the appropriate modelling approaches based on the properties of the data and the feasibility of ensemble learning. Overall, three models demonstrate that CO2 emissions are the primary factor influencing economic development, revealing a strong correlation between industrial or energy-related activities and economic performance. Renewable energy may also facilitate sustainable growth, while labour and capital have limited or adverse effects, exhibiting complex dynamics that vary by environment. The findings show that HDE and GBR are the best models for predicting economic growth and pollutant emissions and accurately capturing intricate non-linear interactions. Additionally, HDE, RF, and GBR offer greater insights into nonlinear interactions than MLR, revealing how these factors affect GDP. Green Total Factor Productivity (GTFP) indicates the influx of capital and labour in Asian countries, facilitating rapid development and industrialisation progress through technological innovation and the development of human capital skills. CO2 emissions and renewable energy influence economic growth, ensuring green technological progress through a productivity-driven approach to maximise the significant effects of energy efficiency and renewable energy, and to support GDP growth. An efficient strategy for utilising these factors is essential, leveraging the combined contributions of their qualities. These results underscore the significance of renewable energy in promoting sustainable development and the complex interplay between economic and environmental factors via implementing Sustainable Development Goals 7 and 13, affordable and clean energy (SDG7), and Climate Action (SDG13) to achieve SDG 8 Decent Work and economic growth and other SDGs of the United Nations (UN) agenda 2030.

Yaping Fu, Fuquan Wang, Zhengyuan Li et al.

Abstract Remanufacturing has become a mainstream sustainable manufacturing paradigm for energy conservation and environmental protection. Disassembly and reprocessing operations are two main activities in remanufacturing. This work proposes multiobjective integrated scheduling of disassembly and reprocessing operations considering product structures and random processing time. First, a stochastic programming model is developed to minimize maximum completion time and total tardiness. Second, a reinforcement learning-based multiobjective evolutionary algorithm is devised considering problem-specific knowledge. Three search strategy combinations are formed: crossover and mutation, crossover and key product-based iterated local search, mutation and key product-based iterated local search. At each iteration, a Q-learning method is devised to intelligently choose a combination of premium strategies. A stochastic simulation is incorporated to evaluate the objective values of the searched solutions. Finally, the formulated model and method are compared with an exact solver, CPLEX, and three well-known metaheuristics from the literature on a set of test instances. The results confirm the excellent competitiveness of the developed model and algorithm for solving the considered problem.

Junchen Ye, Yuhao Mao, Ke Cheng et al.

Given the swift proliferation of structural health monitoring (SHM) technology within tunnel engineering, there is a demand on proficiently and precisely imputing the missing monitoring data to uphold the precision of disaster prediction. In contrast to other SHM datasets, the monitoring data specific to tunnel engineering exhibits pronounced spatiotemporal correlations. Nevertheless, most methodologies fail to adequately combine these types of correlations. Hence, the objective of this study is to develop spatiotemporal recurrent neural network (ST-RNN) model, which exploits spatiotemporal information to effectively impute missing data within tunnel monitoring systems. ST-RNN consists of two moduli: a temporal module employing recurrent neural network (RNN) to capture temporal dependencies, and a spatial module employing multilayer perceptron (MLP) to capture spatial correlations. To confirm the efficacy of the model, several commonly utilized methods are chosen as baselines for conducting comparative analyses. Furthermore, parametric validity experiments are conducted to illustrate the efficacy of the parameter selection process. The experimentation is conducted using original raw datasets wherein various degrees of continuous missing data are deliberately introduced. The experimental findings indicate that the ST-RNN model, incorporating both spatiotemporal modules, exhibits superior interpolation performance compared to other baseline methods across varying degrees of missing data. This affirms the reliability of the proposed model.

Alexandra Mazak-Huemer, Christian Huemer, Michael Vierhauser et al.

With the increasing significance of Research, Technology, and Innovation (RTI) policies in recent years, the demand for detailed information about the performance of these sectors has surged. Many of the current tools are limited in their application purpose. To address these issues, we introduce a requirements engineering process to identify stakeholders and elicitate requirements to derive a system architecture, for a web-based interactive and open-access RTI system monitoring tool. Based on several core modules, we introduce a multi-tier software architecture of how such a tool is generally implemented from the perspective of software engineers. A cornerstone of this architecture is the user-facing dashboard module. We describe in detail the requirements for this module and additionally illustrate these requirements with the real example of the Austrian RTI Monitor.

Jelena Jovanovic

Na Li, Yin Dong, Gaofeng Zhang

Introduction: China’s rapidly aging population and rise in chronic diseases put immense strain on the country’s healthcare system. To address these challenges, Yuhuan People’s Hospital established County-level Integrated Health Organization (CIHO) as part of the Healthy China 2030 initiative. Description: Based on the Kaiser Permanente (KP) model, the CIHO takes a multi-disciplinary, collaborative approach to deliver integrated care. It brings together various medical specialties, collaborates with community organizations and companies, and implements reforms in information technology and payment models. Through these efforts, the CIHO has significantly improved healthcare delivery in Yuhuan county. Discussion: Population segmentation relies on data integration and segmentation tools to identify targeted healthcare needs. The allocation and collaboration of health workforce for residents with different health conditions are suggested to be dynamically designed according to both internal and external factors. Corresponding payment mechanism is also an important factor that needs to be taken into consideration. Conclusion: The CIHO’s success has provided a model for integrated, efficient healthcare that could be replicated in other regions of China and offer insights for rural areas in other countries facing similar demographic and epidemiological pressures.

Chunmei Zhang, Aoran Zhang, Li Zhang et al.

Learning high-quality representations of users, items, and tags from historical interactive data is crucial for personalized tag recommendation (PTR) systems. Currently, most PTR models are committed to learning representations from first-order interactions without considering the exploitation of high-order interactive relations, which can be beneficial for avoiding sub-optimal learning. Although several PTR models equipped with graph neural networks (GNN) have been proposed to capture higher-order semantic relevance from raw data, they all carry out representation learning in Euclidean space, which can still easily result in sub-optimal learning due to embedding distortion. In order to further improve the quality of representation learning for PTR, the paper proposes a novel PTR model based on a lightweight GNN framework with hyperbolic embedding, namely GHPTR. GHPTR explicitly injects higher-order relevance into entity representation through the message propagation and aggregation mechanism of GNN and leverages hyperbolic embedding to alleviate the embedding distortion problem. Experimental results on real-world datasets have demonstrated the superiority of our model over its Euclidean counterparts and state-of-the-art baselines.

Jan Karbowski

This paper provides a perspective on applying the concepts of information thermodynamics, developed recently in non-equilibrium statistical physics, to problems in theoretical neuroscience. Historically, information and energy in neuroscience have been treated separately, in contrast to physics approaches, where the relationship of entropy production with heat is a central idea. It is argued here that also in neural systems information and energy can be considered within the same theoretical framework. Starting from basic ideas of thermodynamics and information theory on a classic Brownian particle, it is shown how noisy neural networks can infer its probabilistic motion. The decoding of the particle motion by neurons is performed with some accuracy and it has some energy cost, and both can be determined using information thermodynamics. In a similar fashion, we also discuss how neural networks in the brain can learn the particle velocity, and maintain that information in the weights of plastic synapses from a physical point of view. Generally, it is shown how the framework of stochastic and information thermodynamics can be used practically to study neural inference, learning, and information storing.

Lucila G. Alvarez-Zuzek, Lucio La Cava, Jelena Grujic et al.

Misleading newsletters can shape individuals' perceptions, and pose a threat to societies; as we witnessed by lowering the severity of follow-up stay-at-home orders and burdening a significant challenge to the fight against COVID-19. In this research, we study (mis)information spreading, reanalyzing behavioral data on online sharing, and analyzing decision-making mechanisms using the Drift Diffusion Model (DDM). We find that subjects display an increased instinctive inclination towards sharing misleading news, but rational thinking significantly curbs this reaction, especially for more cautious and older individuals. On top of network structures with similar characteristics as X, Mastodon, and Facebook, we use an agent-based model to expand this individual knowledge to a large scale where individuals are exposed to (mis)information through friends and share (or not) content with probabilities driven by DDM. We found that the natural shape of these social online networks provides a fertile ground for any news to rapidly become viral. Yet we have found that, for the case of X, limiting the number of followers of the most connected users proves to be an appropriate and feasible containment strategy.

Weronika Łajewska, Krisztian Balog

Generative AI models face the challenge of hallucinations that can undermine users' trust in such systems. We approach the problem of conversational information seeking as a two-step process, where relevant passages in a corpus are identified first and then summarized into a final system response. This way we can automatically assess if the answer to the user's question is present in the corpus. Specifically, our proposed method employs a sentence-level classifier to detect if the answer is present, then aggregates these predictions on the passage level, and eventually across the top-ranked passages to arrive at a final answerability estimate. For training and evaluation, we develop a dataset based on the TREC CAsT benchmark that includes answerability labels on the sentence, passage, and ranking levels. We demonstrate that our proposed method represents a strong baseline and outperforms a state-of-the-art LLM on the answerability prediction task.

Jiahui Wei, Philippe Mary, Elsa Dupraz

In the context of goal-oriented communications, this paper addresses the achievable rate versus generalization error region of a learning task applied on compressed data. The study focuses on the distributed setup where a source is compressed and transmitted through a noiseless channel to a receiver performing polynomial regression, aided by side information available at the decoder. The paper provides the asymptotic rate generalization error region, and extends the analysis to the non-asymptotic regime.Additionally, it investigates the asymptotic trade-off between polynomial regression and data reconstruction under communication constraints. The proposed achievable scheme is shown to achieve the minimum generalization error as well as the optimal rate-distortion region.

Gokul Gowri, Xiao-Kang Lun, Allon M. Klein et al.

Mutual information (MI) is a general measure of statistical dependence with widespread application across the sciences. However, estimating MI between multi-dimensional variables is challenging because the number of samples necessary to converge to an accurate estimate scales unfavorably with dimensionality. In practice, existing techniques can reliably estimate MI in up to tens of dimensions, but fail in higher dimensions, where sufficient sample sizes are infeasible. Here, we explore the idea that underlying low-dimensional structure in high-dimensional data can be exploited to faithfully approximate MI in high-dimensional settings with realistic sample sizes. We develop a method that we call latent MI (LMI) approximation, which applies a nonparametric MI estimator to low-dimensional representations learned by a simple, theoretically-motivated model architecture. Using several benchmarks, we show that unlike existing techniques, LMI can approximate MI well for variables with $> 10^3$ dimensions if their dependence structure has low intrinsic dimensionality. Finally, we showcase LMI on two open problems in biology. First, we approximate MI between protein language model (pLM) representations of interacting proteins, and find that pLMs encode non-trivial information about protein-protein interactions. Second, we quantify cell fate information contained in single-cell RNA-seq (scRNA-seq) measurements of hematopoietic stem cells, and find a sharp transition during neutrophil differentiation when fate information captured by scRNA-seq increases dramatically.

Appel Mahmud, Zulqurnain Ali, Md Ashanuzzaman et al.

The psychology of micro-corporate social responsibility (micro-CSR) and employee outcome has emerged in the contemporary literature of interdisciplinary management science. Previous studies have ignored the testing of mediating effects and boundary conditions in the association between micro-CSR and employee outcomes. Drawing on social identity theory (SIT) and social information processing theory (SIPT), this research aims to investigate how, why, and when the perceived CSR community (PCSRc; a micro-CSR activity) affects employee societal behavior (ESB; a voluntary behavior) accounting the mediating role of perceived external prestige (PEP) and moderating role of organizational identification (OI). Our research recruited 452 employees in Bangladesh via questionnaire and tested the proposed measurement model and structural relationships in AMOS. The results report a significant and positive relationship between PCSRc and ESB. It also reveals that PEP mediates PCSRc and ESB link, and OI regulates the straight association of PCSRc and PEP and ancillary links of PCSRc and ESB (via PEP). Finally, we recorded the research implications and future research directions.

Ehsan Inam Ullah, Shakil Ahmad, Muhammad Fahim Khokhar et al.

Run off river schemes are getting widespread importance as they are considered environmentally safe. However, number of studies and the consequent information regarding impacts of run off river schemes is very limited worldwide. Present study attempted to analyze impacts of Ghazi Barotha Hydropower Plant, which is a run off river scheme situated in Khyber Pakhtunkhwa province of Pakistan. This study attempted to analyze impacts of this run off river scheme on hydrological and ecological conditions of downstream areas. Data on river discharge, groundwater levels, agriculture area, vegetation and bare soil was utilized for this study. All data sets between the year 1990 till 2020 were analyzed. Hydrological impacts were analyzed through secondary data analysis, whereas ecological impacts were studied through remote sensing technique. Statistical methods were applied to further draw conclusions between hydrological and ecological interrelationships. Results showed that after functioning of Ghazi Barotha, there was 47% and 91% reduction of river discharge, in summer and winter seasons respectively. Groundwater level dropped by 50%. Agriculture area reduced by 1.69% and 9.11% during summer and winter respectively, whereas land under bare soil increased. River water diversion was considered to be responsible for groundwater reduction, as strong correlation was found between both. Agriculture land recovery, in post Ghazi Barotha period, was premised at intense groundwater mining, as groundwater level and agriculture area were significantly related (p < 0.05). Governments’ groundwater development schemes, and a shift into motorized groundwater mining were major factors behind further groundwater exploitation in study area. This study came to the conclusion that Ghazi Barotha Hydropower Plant had impacted flow regime of Indus River, as well as groundwater levels and land use of downstream area along the river. These effects were triggered by inappropriate compensatory measures and uncontrolled water resource exploitation.

Sohee Park, Young-Kyun Kwon, Mina Yoon et al.

Abstract In many complex oxides, the oxygen vacancy formation is a promising route to modify the material properties such as a superconductivity and an oxygen diffusivity. Cation substitutions and external strain have been utilized to control the concentration and diffusion of oxygen vacancies, but the mechanisms behind the controls are not fully understood. Using first-principles calculations, we find how Sr doping and external strain greatly enhances the diffusivity of oxygen vacancies in La2−x Sr x CuO4−δ (LSCO) in the atomic level. In hole-doped case (2x > δ), the formation energy of an apical vacancy in the LaO layer is larger than its equatorial counterpart by 0.2 eV that the bottleneck of diffusion process is for oxygen vacancies to escape equatorial sites. Such an energy difference can be reduced and even reversed by either small strain (< 1.5%) or short-range attraction between Sr and oxygen vacancy, and in turn, the oxygen diffusivity is greatly enhanced. For fully compensated hole case (2x ≦ δ), the formation energy of an apical vacancy becomes too high that most oxygen vacancies cannot move but would be trapped at equatorial sites. From our electronic structure analysis, we found that the contrasting change in the formation energy by Sr doping and external strain is originated from the different localization natures of electron carrier from both types of oxygen vacancies.

Amrita Kumari, Neha Raina, Abhishek Wahi et al.

Wound healing is an intricate process of tissue repair or remodeling that occurs in response to injury. Plants and plant-derived bioactive constituents are well explored in the treatment of various types of wounds. Curcumin is a natural polyphenolic substance that has been used since ancient times in Ayurveda for its healing properties, as it reduces inflammation and acts on several healing stages. Several research studies for curcumin delivery at the wound site reported the effectiveness of curcumin in eradicating reactive oxygen species and its ability to enhance the deposition of collagen, granulation tissue formation, and finally, expedite wound contraction. Curcumin has been widely investigated for its wound healing potential but its lower solubility and rapid metabolism, in addition to its shorter plasma half-life, have limited its applications in wound healing. As nanotechnology has proven to be an effective technique to accelerate wound healing by stimulating appropriate mobility through various healing phases, curcumin-loaded nanocarriers are used for targeted delivery at the wound sites. This review highlights the potential of curcumin and its nanoformulations, such as liposomes, nanoparticles, and nano-emulsions, etc. in wound healing. This paper emphasizes the numerous biomedical applications of curcumin which collectively prepare a base for its antibiofilm and wound-healing action.

Ze‐Jiaxin Niu, Shi Yao, Xi Zhang et al.

Abstract Neural precursor cell expressed developmentally downregulated 4‐like (NEDD4L), a member of the E3 ubiquitin‐protein ligases, encoded by NEDD4L gene, was found to be involved in in salt sensitivity by regulating sodium reabsorption in salt‐sensitive rats. The authors aimed to explore the associations of NEDD4L genetic variants with salt sensitivity, blood pressure (BP) changes and hypertension incidence in Chinese adults. Participants from 124 families in Northern China in the Baoji Salt‐Sensitive Study Cohort in 2004, who received the chronic salt intake intervention, including a 7‐day low‐salt diet (3.0 g/day) and a 7‐day high‐salt diet (18 g/day), were analyzed. Besides, the development of hypertension over 14 years was evaluated. NEDD4L single nucleotide polymorphism (SNP) rs74408486 was shown to be significantly associated with systolic BP (SBP), diastolic BP (DBP) and mean arterial pressure (MAP) responses to low‐salt diet, while SNPs rs292449 and rs2288775 were significantly associated with pulse pressure (PP) response to high‐salt diet. In addition, SNP rs4149605, rs73450471, and rs482805 were significantly associated with the longitudinal changes in SBP, DBP, MAP, or PP at 14 years of follow‐up. SNP rs292449 was significantly associated with hypertension incidence over the 14‐year follow‐up. Finally, this gene‐based analysis found that NEDD4L was significantly associated with longitudinal BP changes and the incidence of hypertension over the 14‐year follow‐up. This study indicated that gene polymorphism in NEDD4L serve an important function in salt sensitivity, longitudinal BP change and development of hypertension in the Chinese population.

Yu-Sun Liu, Chun-Hung Huang, Shingchern D. You

In this paper, we propose a method to estimate the mean square error (MSE) of the estimated channel for ATSC (Advanced Television Systems Committee) 3.0 systems. When combining the channel MSE and noise variance, we can better estimate the a priori LLR (log likelihood ratio) for the sum–product algorithm. The experimental results show that doing so yields better BER (bit error rate) performance in the 0 dB echo channel. The improvement in the 2-D channel estimation is about 0.2 dB. In the 1-D estimation case, the proposed approach is essential to decode codewords.