Search Results (861)

Crowdfunding platforms are important for startups, since they offer diverse financing options, market validation, and promotional opportunities through an investor community. These platforms provide detailed company information, aiding informed investment decisions within a regulated and secure environment. Machine learning (ML) techniques are important in analyzing large data sets, detecting anomalies and fraud, and enhancing decision-making and business strategies. A systematic review employed PRISMA guidelines, which studied how ML improves fraud detection on digital crowdfunding platforms. The analysis includes English-language studies from peer-reviewed journals published between 2018 and 2023 to analyze the pre- and post-COVID-19 pandemic. The findings indicate that ML techniques such as Random Forest, Support Vector Machine, and Artificial Neural Networks significantly enhance the predictive accuracy and utility of tax planning for startups considering equity crowdfunding. The United States, Germany, Canada, Italy, and Turkey do not present statistically significant differences at the 95% confidence level, standing out for their notable academic visibility. Florida Atlantic and Cornell Universities, Springer and John Wiley & Sons Ltd. publishing houses, and the Journal of Business Ethics and Management Science magazines present the highest citations without statistical differences at the 95% confidence level. Full article

In the cybersecurity industry, where legitimate transactions far outnumber fraudulent ones, detecting fraud is of paramount significance. In order to evaluate the accuracy of detecting fraudulent transactions in imbalanced real datasets, this study compares the efficacy of two approaches, random under-sampling and oversampling, using the synthetic minority over-sampling technique (SMOTE). Random under-sampling aims for fairness by excluding examples from the majority class, but this compromises precision in favor of recall. To strike a balance and ensure statistical significance, SMOTE was used instead to produce artificial examples of the minority class. Based on the data obtained, it is clear that random under-sampling achieves high recall (92.86%) at the expense of low precision, whereas SMOTE achieves a higher accuracy (86.75%) and a more even F1 score (73.47%) at the expense of a slightly lower recall. As true fraudulent transactions require at least two methods for verification, we investigated different machine learning methods and made suitable balances between accuracy, F1 score, and recall. Our comparison sheds light on the subtleties and ramifications of each approach, allowing professionals in the field of cybersecurity to better choose the approach that best meets the needs of their own firm. This research highlights the need to resolve class imbalances for effective fraud detection in cybersecurity, as well as the need for constant monitoring and the investigation of new approaches to increase applicability. Full article

Credit card fraud detection is a critical issue for financial institutions due to significant financial losses and the erosion of customer trust. Fraud not only impacts the bottom line but also undermines the confidence customers place in financial services, leading to long-term reputational damage. Traditional machine learning methods struggle to improve detection accuracy with limited data, adapt to new fraud techniques, and detect complex fraud patterns. To address these challenges, we present FedGAT-DCNN, a model integrating a Graph Attention Network (GAT) and dilated convolutions within a federated learning framework. FedGAT-DCNN employs federated learning, allowing financial institutions to collaboratively train models using local datasets, enhancing accuracy and robustness while maintaining data privacy. Incorporating a GAT enables continuous model updates across institutions, quickly adapting to new fraud patterns. Dilated convolutions extend the model’s receptive field without extra computational overhead, improving detection of subtle and complex fraudulent activities. Experiments on the 2018CN and 2023EU datasets show that FedGAT-DCNN outperforms traditional models and other federated learning methods, achieving a ROC-AUC of 0.9712 on the 2018CN dataset and 0.9992 on the 2023EU dataset. These results highlight FedGAT-DCNN’s robustness, accuracy, and applicability in real-world fraud detection scenarios. Full article

Corporate fraud poses a significant obstacle for sustainable business development. Drawing on social network analysis, this paper used data originated from Chinese-listed companies from 2009 to 2022 and found that directors’ network position significantly mitigates corporate fraud. Mechanism tests indicated that the quality of external auditors and internal control play a mediating role in this relationship. Further analysis showed that the network positions of independent directors, non-independent directors, and female directors individually inhibit the inclination of corporate fraud when considering various types of directors. Of note, the busy director hypothesis was not applicable in explaining the impact of directors’ network position on corporate fraud. This study provides a new approach to improving the sustainability of enterprises in newly emerging markets via the analysis of director networks. It is also beneficial to the research on director networks and corporate fraud in companies, offering insights for corporate governance and fraud prevention in companies and regulatory agencies. Full article

Machine learning models play a critical role in applications such as image recognition, natural language processing, and medical diagnosis, where accuracy and efficiency are paramount. As datasets grow in complexity, so too do the computational demands of classification techniques. Previous research has achieved high accuracy but required significant computational time. This paper proposes a parallel architecture for Ensemble Machine Learning Models, harnessing multicore CPUs to expedite performance. The primary objective is to enhance machine learning efficiency without compromising accuracy through parallel computing. This study focuses on benchmark ensemble models including Random Forest, XGBoost, ADABoost, and K Nearest Neighbors. These models are applied to tasks such as wine quality classification and fraud detection in credit card transactions. The results demonstrate that, compared to single-core processing, machine learning tasks run 1.7 times and 3.8 times faster for small and large datasets on quad-core CPUs, respectively. Full article

Food adulteration which is economically motivated (i.e., food fraud) is an incentive for the development and application of new and fast detection methods/instruments. An example of a fast method that is extremely environmentally friendly is near-infrared spectroscopy (NIRS). Therefore, the goal of this research was to examine the potential of its application in monitoring the adulteration of blended sunflower/olive oils and to compare two types of NIRS instruments, one of which is a portable micro-device, which could be used to assess the purity of olive oil anywhere and would be extremely useful to inspection services. Both NIR devices (benchtop and portable) enable absorbance monitoring in the wavelength range from 900 to 1700 nm. Extra virgin oils (EVOOs) and “ordinary” olive oils (OOs) from large and small producers were investigated, which were diluted with sunflower oil in proportions of 1–15%. However, with the appearance of different salad oils that have a defined share of EVOO stated on the label (usually 10%), the possibilities of the recognition and manipulation in these proportions were tested; therefore, EVOO was also added to sunflower oil in proportions of 1–15%. The composition of fatty acids, color parameters, and total dissolved substances and conductivity for pure and “adulterated” oils were monitored. Standard tools of multivariate analysis were applied, such as (i) analysis of main components for the qualitative classification of oil and (ii) partial regression using the least square method for quantitative prediction of the proportion of impurities and fatty acids. Qualitative models proved successful in classifying (100%) the investigated oils, regardless of whether the added thinner was olive or sunflower oil. Developed quantitative models relating measured parameters with the NIR scans, resulted in values of R² ≥ 0.95 and was reliable (RPD > 8) for fatty acid composition prediction and for predicting the percentage of the added share of impurity oils, while color attributes were less successfully predicted with the portable NIR device (RPD in the range of 2–4.2). Although with the portable device, the prediction potentials remained at a qualitative level (e.g., color parameters), it is important to emphasize that both devices were tested not only with EVOO but also with OO and regardless of whether proportions of 1–15% sunflower oil were added to EVOO and OO or EVOO and OO in the same proportions to sunflower oil. Full article

Fraudulent and adulterated food is produced mainly to reduce prices and attract consumers’ attention whilst threatening their economic interests, health, and safety. As such, this type of activity should be eliminated. This study’s aim was to identify the most common hazards related to food fraud and adulteration, reported in the Rapid Alert System for Food and Feed (RASFF) between 2000 and 2021, taking into account the product category (including individual products), country of origin, and notification type. We used Microsoft Excel (filtering, vertical-searching, transposition, and pivot table functions) and Statistica 13.3 (two-way joining cluster analysis) to analyse similarities between the hazards identified throughout the research period. Notifications relating to food fraud and adulteration accounted for 18.7% of all RASFF notifications, fluctuating between 1000 and 1200 per year in recent years. These mainly included hazards related to composition and novel foods in dietetic foods, food supplements and fortified foods, sulphites in fruits and vegetables, colours in cereals and bakery products, or Sudan in herbs and spices. Dietetic foods, dietary supplements, and fortified foods were mainly reported as alerts and information notifications, meaning that they were already available on the common European internal market. The other products originated mainly from outside the European Union (Asia—Turkey, Uzbekistan, India and Africa—Ghana, Nigeria) and were submitted on the basis of border rejections. Therefore, it is necessary to continue closely monitoring imported products at the EU border to ensure food safety, avoiding fraud and adulteration and protecting consumers’ financial interests. Full article

The adsorption of p-aminothiophenol (PATP) on metallic nanostructures is a very interesting phenomenon that depends on many factors, and because of that, PATP is an increasingly important probe molecule in surface-enhanced Raman spectroscopy (SERS) due to its strong interaction with Ag and Au, its intense SERS signal, and its significance in molecular electronics. In our study, the SERS spectra of PATP on silver colloids were investigated and we considered several factors, such as the effect of the adsorbate concentration, the nature of the metallic nanoparticles, and the excitation wavelength. Differences between the SERS spectra recorded at high and low concentrations of PATP were explained and DFT calculations of different species were performed in order to support the experimental results. Additionally, time-dependent density-functional theory (TD-DFT) calculations were used to simulate the UV spectra of each species and to determine the MOs involved in each transition. The presence of different species of PATP adsorbed onto the metal surface gave rise to the acquisition of simultaneous SERS signals from those species and the consequent overlapping of some bands with new SERS bands coming from the dimerization of PATP. This work helped to discern which species is responsible for each SERS spectrum under particular experimental conditions. Full article

Given the recognized nutritional value of fish and shifting consumer lifestyles, processed seafood has become increasingly prevalent, comprising a significant portion of global food production. Although current European Union labeling regulations do not require species declaration for these products, food business operators often voluntarily provide this information on ingredient lists. Next Generation Sequencing (NGS) approaches are currently the most effective methods for verifying the accuracy of species declarations on processed seafood labels. This study examined the species composition of 20 processed seafood products, each labeled as containing a single species, using two DNA metabarcoding markers targeting the mitochondrial cytochrome c oxidase I (COI) and 16S rRNA genes. The combined use of these markers revealed that the majority of the products contained multiple species. Furthermore, two products were found to be mislabeled, as the declared species were not detected. These findings underscore that NGS is a robust technique that could be adopted to support routine food industry activities and official control programs, thereby enhancing the ‘From Boat to Plate’ strategy and combating fraudulent practices in the complex fisheries supply chain. Full article

We examine the impact of corporate governance and personal networks on corporate fraud in Japanese companies, using panel logit and Cox proportional hazard models to analyze fraud occurrence and detection. This study focuses on the effects of Japan’s recent corporate governance reform and explores the unique influence of personal networks. Our key findings indicate that recent changes in corporate governance in Japan have been effective in preventing the occurrence of fraud and accelerating its detection. Additionally, stronger personal networks among board members help prevent fraud concealment, highlighting cultural differences in the effectiveness of personal networks in corporate governance compared to findings from Europe and the US. Full article

This study addresses the critical challenge of data imbalance in credit card fraud detection (CCFD), a significant impediment to accurate and reliable fraud prediction models. Fraud detection (FD) is a complex problem due to the constantly evolving tactics of fraudsters and the rarity of fraudulent transactions compared to legitimate ones. Efficiently detecting fraud is crucial to minimize financial losses and ensure secure transactions. By developing a framework that transitions from imbalanced to balanced data, the research enhances the performance and reliability of FD mechanisms. The strategic application of Meta-heuristic optimization (MHO) techniques was accomplished by analyzing a dataset from Kaggle’s CCF benchmark datasets, which included data from European credit-cardholders. They evaluated their capability to pinpoint the smallest, most relevant set of features, analyzing their impact on prediction accuracy, fitness values, number of selected features, and computational time. The study evaluates the effectiveness of 15 MHO techniques, utilizing 9 transfer functions (TFs) that identify the most relevant subset of features for fraud prediction. Two machine learning (ML) classifiers, random forest (RF) and support vector machine (SVM), are used to evaluate the impact of the chosen features on predictive accuracy. The result indicated a substantial improvement in model efficiency, achieving a classification accuracy of up to 97% and reducing the feature size by up to 90%. In addition, it underscored the critical role of feature selection in optimizing fraud detection systems (FDSs) and adapting to the challenges posed by data imbalance. Additionally, this research highlights how machine learning continues to evolve, revolutionizing FDSs with innovative solutions that deliver significantly enhanced capabilities. Full article

From the various perspectives of machine learning (ML) and the multiple models used in this discipline, there is an approach aimed at training models for the early detection (ED) of anomalies. The early detection of anomalies is crucial in multiple areas of knowledge since identifying and classifying them allows for early decision making and provides a better response to mitigate the negative effects caused by late detection in any system. This article presents a literature review to examine which machine learning models (MLMs) operate with a focus on ED in a multidisciplinary manner and, specifically, how these models work in the field of fraud detection. A variety of models were found, including Logistic Regression (LR), Support Vector Machines (SVMs), decision trees (DTs), Random Forests (RFs), naive Bayesian classifier (NB), K-Nearest Neighbors (KNNs), artificial neural networks (ANNs), and Extreme Gradient Boosting (XGB), among others. It was identified that MLMs operate as isolated models, categorized in this article as Single Base Models (SBMs) and Stacking Ensemble Models (SEMs). It was identified that MLMs for ED in multiple areas under SBMs’ and SEMs’ implementation achieved accuracies greater than 80% and 90%, respectively. In fraud detection, accuracies greater than 90% were reported by the authors. The article concludes that MLMs for ED in multiple applications, including fraud, offer a viable way to identify and classify anomalies robustly, with a high degree of accuracy and precision. MLMs for ED in fraud are useful as they can quickly process large amounts of data to detect and classify suspicious transactions or activities, helping to prevent financial losses. Full article

Element profiling is a powerful tool for detecting fraud related to claims of geographical origin. However, these methods must be continuously developed, as mixtures of different origins in particular offer great potential for adulteration. This study is a proof of principle to determine whether elemental profiling is suitable for detecting mixtures of the same food but from different origins and whether calculated data from walnut mixtures could help to reduce the measurement burden. The calculated data used in this study were generated based on measurements of authentic, unadulterated samples. Five different classification models and three regression models were applied in five different evaluation approaches to detect adulteration or even distinguish between adulteration levels (10% to 90%). To validate the method, 270 mixtures of walnuts from different origins were analyzed using inductively coupled plasma mass spectrometry (ICP-MS). Depending on the evaluation approach, different characteristics were observed in mixtures when comparing the calculated and measured data. Based on the measured data, it was possible to detect admixtures with an accuracy of 100%, even at low levels of adulteration (20%), depending on the country. However, calculated data can only contribute to the detection of adulterated walnut samples in exceptional cases. Full article

Extra virgin olive oil (EVOO) is a precious and healthy ingredient of Mediterranean cuisine. Due to its high nutritional value, the interest of consumers in the composition of EVOO is constantly increasing, making it a product particularly exposed to fraud. Therefore, there is a need to properly valorize high-quality EVOO and protect it from fraudulent manipulations to safeguard consumer choices. In our study, we used a straightforward and easy method to assess the molecular traceability of 28 commercial EVOO samples based on the use of SSR molecular markers. A lack of correspondence between the declared origin of the samples and the actual origin of the detected varieties was observed, suggesting possible adulteration. This result was supported by the identification of private alleles based on a large collection of national and international olive varieties and the search for them in the molecular profile of the analyzed samples. We demonstrated that the proposed method is a rapid and straightforward approach for identifying the composition of an oil sample and verifying the correspondence between the origin of olives declared on the label and that of the actual detected varieties, allowing the detection of possible adulterations. Full article

Given its huge economic, nutritional, and social value, extra virgin olive oil (EVOO) is an essential food. This flagship product of the countries bordering the Mediterranean basin is one of the most frauded products worldwide. Although traditional chemical analyses have demonstrated to be reliable tools for olive drupes and EVOO quality assessment, they present several drawbacks; the urgent need for fast and non-destructive techniques thus motivated this review. Given the lack of comprehensive reviews in the literature, our first aim was to summarize the current knowledge regarding applying spectroscopies, imaging analyses, and other non-destructive techniques to olives and EVOO. The second aim was to highlight the most innovative and futuristic applications and outline the future research prospects within this strategic production chain. With respect to olive drupes, the most interesting results were obtained using RGB imaging and NIR spectroscopy, particularly using portable NIR devices and specific digital cameras for in-field or in-mill monitoring. Nevertheless, it is important to highlight that RGB imaging and NIR spectroscopy need to be integrated with flesh hardness measurements, given the higher reliability of this parameter compared to olive skin color. Finally, with respect to EVOO, although several useful applications of visible imagining, UV–Visible, NIR, and Mid-Infrared spectroscopies have been found, the online monitoring of EVOO quality using NIR spectroscopy strikes us as being the most interesting technique for improving the EVOO production chain in the near future. Full article