Search Results (9)

Honey variety authentication is critical for ensuring market integrity and protecting consumer rights, especially for high-value unifloral honeys, such as acacia honey, which are frequently adulterated with low-value alternatives such as rape honey due to their similar visual appearance. The aim of this study was to develop a method for precise discrimination between rape honey and acacia honey using their chemical profiles combined with machine learning. A total of 542 honey samples were collected from major beekeeping regions in China. Targeted quantification of 12 sugars and 20 amino acids was performed using UPLC-MS/MS. Multivariate analysis revealed significant differences in sugar and amino acid compositions between the two honey types, though partial samples overlapped due to chemical similarity. Six machine learning algorithms, including the Multilayer Perceptron, were employed for classification. Optimization was performed via 10-fold cross-validation and ADASYN oversampling, yielding optimal performance of 98% and 100% prediction accuracies for rape honey and acacia honey, respectively, on the independent test set. SHAP (Shapley Additive Explanations) analysis identified key differential markers, including fructose, turanose, glucose, and GABA, which contributed most to the classification. Furthermore, a user-friendly web application was developed to facilitate rapid on-site authentication. This study provides an innovative technical framework for honey variety discrimination, with potential applications in quality control and anti-fraud practices. Full article

The relationships within a supply chain are crucial for analyzing business transactions and can reveal significant patterns in disclosed financial data. These relationships also aid in the assessment and detection of financial fraud. Recent studies employing graph neural networks (GNNs) have demonstrated enhanced detection capabilities by integrating corporate financial features with supply chain relationships, surpassing traditional methods that rely solely on financial features. However, these studies face notable limitations: (1) they do not model enterprise associations across consecutive years, hindering the detection of long-term financial fraud, and (2) they lack efficacy in interpretive analyses of supply chain relationships to uncover patterns of fraud or risk transfer. To address these gaps, this paper introduces an interpretable and efficient Heterogeneous Graph Convolutional Network (ieHGCN) designed to analyze supply chain knowledge graphs. It also extends the model’s learning scope to multi-year financial data for detecting fraud. The experimental results indicate that our method, offering both extensibility and interpretability, significantly outperforms existing machine learning and GNN approaches in continuous multi-year fraud detection, achieving the highest AUC of 0.7498, a 3.8% improvement over the leading method. Furthermore, meta-path analysis reveals that companies sharing the same supplier exhibit correlated financial fraud risks and that fraud can propagate through the supply chain, providing novel insights into anti-fraud and risk management strategies through enhanced interpretability. Full article

Fraud detection within transaction data is crucial for maintaining financial security, especially in the era of big data. This paper introduces a novel fraud detection method that utilizes quantum computing to implement community detection in transaction networks. We model transaction data as an undirected graph, where nodes represent accounts and edges indicate transactions between them. A modularity function is defined to measure the community structure of the graph. By optimizing this function through the Quadratic Unconstrained Binary Optimization (QUBO) model, we identify the optimal community structure, which is then used to assess the fraud risk within each community. Using a Coherent Ising Machine (CIM) to solve the QUBO model, we successfully divide 308 nodes into four communities. We find that the CIM computes faster than the classical Louvain and simulated annealing (SA) algorithms. Moreover, the CIM achieves better community structure than Louvain and SA as quantified by the modularity function. The structure also unambiguously identifies a high-risk community, which contains almost 70% of all the fraudulent accounts, demonstrating the practical utility of the method for banks’ anti-fraud business. Full article

The proliferation of internet services in various industries, especially the financial sector, has increased financial fraud. Fraud detection and prevention are critical to protecting both individuals and organizations from significant financial loss. However, the lack of publicly available datasets containing fraud is a major challenge. This study aims to address these issues using advanced machine learning techniques. Known for their ability to provide insight into data, decision trees are used for real-time fraud detection. In addition, deep learning techniques and artificial neural networks (ANN) are used to detect complex fraud patterns, while logistic regression is used to model the probability of fraudulent events. The accuracy of these methods, including decision trees, logistic regression, and ANN, is fully evaluated, with accuracies of 99.8%, 99.9%, and 99.94%, respectively. These findings provide valuable guidance for companies on choosing effective anti-fraud strategies and shed light on the adaptability of algorithms to real financial contexts, contributing to machine learning-based fraud detection. Full article

The road surfacing material is continuously exposed to both traffic and climate stresses, including extreme temperatures and heavy rainfalls. Subsequently, most road defects start from the surface and potentially develop into road structural damage. Appropriate road maintenance schemes are necessary to ensure the surface remains functionable. On the other hand, these maintenance schemes would be uneconomical without the use of durable and climate-resilient materials for road surfacing. This is vital in low-income countries (LICs), where the burdens of road maintenance are economically unbearable. In this regard, long-life epoxy-modified asphalt offers an opportunity to achieve durable and climate-resilient pavement surfacing. However, epoxy is an expensive material that may be subject to fraud, leading to poor quality of the surfacing mixture and the resulting road infrastructure. In order to prevent fraud and ensure the quality of epoxy bitumen mixtures for road surfacing, Fourier transform infrared spectroscopy (FTIR) was used to characterise mixes of epoxy-modified bitumen and trace the content of epoxy. The findings showed that the epoxy content used in the preparation of mixes agreed with the epoxy content traced using FTIR. The mean difference between the two quantities was approximately ±1.0% with a correlation coefficient R² > 0.9. Therefore, FTIR can efficiently provide an antifraud method for modified epoxy asphalt mixes at the plant level to help achieve sustainable pavements. Full article

Empirical research identifies whistleblowing as one of the most effective internal antifraud controls. Very recently, Directive 1937/2019 became effective in the EU, aiming to deal with the defragmentation of whistleblowing legislation among the member states and provide common minimum accepted standards. The present article aims to provide a verified, weighted comparative maturity model. The suggested model has been constructed based on the methodology for constructing comparative maturity models and validated based on the Delphi method. The weights on each validated component have been calculated based on the summing of votes method. The study resulted in eight main components «scope», «corporate governance», «reporting mechanisms», «protection», «tone at the top», «organizational and human resource practices», «investigations» and «monitor and review» divided further into 18 elements. The suggested maturity model may provide a pathway for organizations to develop and maintain a robust whistleblowing maturity framework that will benefit both the organizations and the public welfare. Full article

The number of fraud occurrences in electronic banking is rising each year. Experts in the field of cybercrime are continuously monitoring and verifying network infrastructure and transaction systems. Dedicated threat response teams (CSIRTs) are used by organizations to ensure security and stop cyber attacks. Financial institutions are well aware of this and have increased funding for CSIRTs and antifraud software. If the company has a rule-based antifraud system, the CSIRT can examine fraud cases and create rules to counter the threat. If not, they can attempt to analyze Internet traffic down to the packet level and look for anomalies before adding network rules to proxy or firewall servers to mitigate the threat. However, this does not always solve the issues, because transactions occasionally receive a “gray” rating. Nevertheless, the bank is unable to approve every gray transaction because the number of call center employees is insufficient to make this possible. In this study, we designed a machine-learning-based rating system that provides early warnings against financial fraud. We present the system architecture together with the new ML-based scoring extension, which examines customer logins from the banking transaction system. The suggested method enhances the organization’s rule-based fraud prevention system. Because they occur immediately after the client identification and authorization process, the system can quickly identify gray operations. The suggested method reduces the amount of successful fraud and improves call center queue administration. Full article

Fourier-Transform mid-infrared (FTIR) spectroscopy offers a strong candidate screening tool for rapid, non-destructive and early detection of unauthorized virgin olive oil blends with other edible oils. Potential applications to the official anti-fraud control are supported by dozens of research articles with a “proof-of-concept” study approach through different chemometric workflows for comprehensive spectral analysis. It may also assist non-targeted authenticity testing, an emerging goal for modern food fraud inspection systems. Hence, FTIR-based methods need to be standardized and validated to be accepted by the olive industry and official regulators. Thus far, several literature reviews evaluated the competence of FTIR standalone or compared with other vibrational techniques only in view of the chemometric methodology, regardless of the inherent characteristics of the product spectra or the application scope. Regarding authenticity testing, every step of the methodology workflow, and not only the post-acquisition steps, need thorough validation. In this context, the present review investigates the progress in the research methodology on FTIR-based detection of virgin olive oil adulteration over a period of more than 25 years with the aim to capture the trends, identify gaps or misuses in the existing literature and highlight intriguing topics for future studies. An extensive search in Scopus, Web of Science and Google Scholar, combined with bibliometric analysis, helped to extract qualitative and quantitative information from publication sources. Our findings verified that intercomparison of literature results is often impossible; sampling design, FTIR spectral acquisition and performance evaluation are critical methodological issues that need more specific guidance and criteria for application to product authenticity testing. Full article

Background: The present work aims at obtaining an approximate early production estimate of olive orchards used for extra virgin olive oil production by combining image analysis techniques with light drone images acquisition and photogrammetric reconstruction. Methods: In May 2019, an orthophoto was reconstructed through a flight over an olive grove to predict oil production from segmentation of plant canopy surfaces. The orchard was divided into four plots (three considered as training plots and one considered as a test plot). For each olive tree of the considered plot, the leaf surface was assessed by segmenting the orthophoto and counting the pixels belonging to the canopy. At harvesting, the olive production per plant was measured. The canopy radius of the plant (R) was automatically obtained from the pixel classification and the measured production was plotted as a function of R. Results: After applying a k-means-classification to the four plots, two distinct subsets emerged in association with the year of loading (high-production) and unloading. For each plot of the training set the logarithm of the production curves against R were fitted with a linear function considering only four samples (two samples belonging to the loading region and two samples belonging to the unloading one) and the total production estimate was obtained by integrating the exponent of the fitting-curve over R. The three fitting curves obtained were used to estimate the total production of the test plot. The resulting estimate of the total production deviates from the real one by less than 12% in training and less than 18% in tests. Conclusions: The early estimation of the total production based on R extracted by the orthophotos can allow the design of an anti-fraud protocol on the declared production. Full article