Sai Srikanth Madugula, Peplluis Esteva De La Rosa, Daya Shankar
The integration of machine learning into decentralized finance (DeFi) credit assessment is frequently undermined by opaque algorithms and severe methodological flaws regarding data leakage. This paper presents a rigorous, fully reproducible framework for explainable artificial intelligence (XAI) in invoice-backed default risk modeling. Utilizing a highly imbalanced dataset of 12,000 corporate loan originations, we engineer an XGBoost ensemble model that achieves an AUC-ROC of 0.89. We systematically eliminate the pervasive data leakage associated with the Synthetic Minority Over-sampling Technique (SMOTE) by implementing a dynamic crossvalidation pipeline, ensuring synthetic data generation is strictly isolated to training folds. To satisfy institutional accounting standards for expected loss (e.g., IFRS 9), we mathematically formulate and validate the Expected Calibration Error (ECE), achieving a highly calibrated probabilistic output of 0.08. Furthermore, we extract local explanations using SHAP (SHapley Additive exPlanations), imposing strict constraints on the background reference dataset to guarantee mathematical additivity and prevent stochastic approximation transitions. Our findings reveal that Days Payment Outstanding (DPO) and invoice age are primary default drivers, while on-chain reputation effectively mitigates perceived risk. Finally, we address critical privacy vulnerabilities, mathematically modeling Membership Inference Attacks (MIAs) on synthetic records. This work establishes a regulatory-compliant, structurally sound ML foundation for permissionless credit provision.
Predictive Churn Modeling and Proactive Service Using Customer Interaction Data Objectives: 1. To provide a comprehensive understanding of cloud-native architectures and middleware technologies used for designing scalable, resilient, and high-performance financial trading systems. 2. To explain the core concepts of microservices, containerization, orchestration, distributed messaging, and data management that power modern financial platforms and digital banking ecosystems. 3. To demonstrate the practical implementation of advanced technologies such as Kubernetes, Apache Kafka, Redis, gRPC, and AI-driven solutions for real-time trading and financial service delivery. 4. To equip software engineers, solution architects, researchers, and FinTech professionals with the knowledge required to build secure, fault-tolerant, low-latency, and highly observable trading infrastructures. 5. To explore emerging trends in financial technology, including serverless computing, WebAssembly, Artificial Intelligence, Machine Learning, and Decentralized Finance (DeFi), preparing readers for the next generation of cloud-native financial systems. Table of Contents CHAPTER 1 The Foundation of Customer Retention: Concepts and Definitions CHAPTER 2 The Business Value of Predicting Churn: Impact on ROI CHAPTER 3 Sources of Customer Interaction Data: CRM, Logs, and Beyond CHAPTER 4 The Architecture of a Churn Prediction System CHAPTER 5 Data Acquisition and Quality Assessment CHAPTER 6 Preprocessing High-Dimensional Interaction Data CHAPTER 7 Feature Engineering: Creating Meaningful Indicators from Raw Data CHAPTER 8 Exploratory Data Analysis for Churn Patterns CHAPTER 9 Traditional Statistical Methods in Churn Modeling CHAPTER 10 Machine Learning Approaches: From Random Forests to XGBoost CHAPTER 11 Deep Learning for Temporal Interaction Sequences CHAPTER 12 Natural Language Processing for Sentiment-Based Churn Analysis CHAPTER 13 Handling Class Imbalance in Churn Datasets CHAPTER 14 Evaluating Model Performance: Beyond Accuracy CHAPTER 15 Interpreting Black-Box Models for Stakeholder Trust CHAPTER 16 Real-Time Churn Scoring and Pipeline Automation CHAPTER 17 Designing Proactive Service Interventions CHAPTER 18 Personalized Marketing and Customer Success Strategies CHAPTER 19 Ethical Considerations and Data Privacy in Churn Modeling CHAPTER 20 Case Studies and Future Trends in Predictive Analytics
Subrat Kumar Jena, Gayatri Palai, Asst. Prof. Rumana Hasinullah Shaikh
Abstract-The rapid expansion of the global gig economy has fundamentally changed the structure of personal finance management. Unlike salaried professionals who operate within predictable monthly income cycles, freelancers and independent contractors face highly volatile cashflow patterns characterized by delayed client payments, irregular project pipelines, seasonal fluctuations, and unstable liquidity reserves. Traditional Personal Financial Management (PFM) systems primarily focus on historical transaction tracking and static budgeting, making them ineffective for proactive financial survival planning in modern freelance ecosystems. This project introduces Prophet AI v1.1, an AI-driven financial intelligence platform engineered specifically to simulate, forecast, and analyze unstable freelance cashflow environments using distributed cloud infrastructure, cryptographic verification, and real-time neural intelligence. The proposed system functions as a Financial Flight Simulator that allows freelancers to model financial risk before it becomes catastrophic in real life. The platform combines machine learning-based forecasting, stochastic risk simulation, cryptographic integrity validation, asynchronous AI orchestration, and multilingual neural voice synthesis within a single integrated ecosystem. The system architecture follows a distributed deployment model consisting of a Next.js 14 frontend hosted on Vercel, a FastAPI Intelligence Gateway hosted on Render, and a Supabase PostgreSQL secure transaction vault. This decoupled architecture ensures scalability, modularity, low frontend latency, and reliable handling of long-running AI inference tasks. The financial forecasting engine utilizes a hybrid intelligence pipeline combining statistical forecasting principles and ensemble-based analytical logic. The platform generates 30-day rolling liquidity forecasts, safe spending corridors, and stress-based runway simulations that help users evaluate financial survival scenarios under varying burn conditions. Unlike conventional financial dashboards, Prophet AI introduces dynamic What-If simulation controls, allowing users to manipulate variables such as liquidity lag, expense escalation, and delayed client payments in real time. To establish institutional-grade trust and forensic-grade auditability, the system implements an Integrity Shield powered by the SHA-256 cryptographic hashing algorithm. Every transaction entered into the system generates a unique digital fingerprint using transaction attributes including amount, date, category, and user identification. This verification mechanism ensures that tampered or manipulated financial records cannot enter the intelligence pipeline, thereby maintaining a Verified Ledger architecture. The project additionally documents real-world deployment challenges involving decimal precision mismatches between JavaScript and Python environments and explains the implementation of strategic normalization bypass mechanisms for stable production deployment. The intelligence layer of Prophet AI is powered using Llama 3.3-70B via Groq infrastructure, enabling high-speed financial reasoning and structured JSON-based strategy generation. The platform utilizes a carefully engineered Ruthless Financial Strategist system prompt designed to deliver direct, survival-oriented financial recommendations rather than emotionally comforting advice. This design philosophy reflects the real-world operational needs of freelancers who require accurate liquidity warnings and actionable strategic insights during financial instability. The generated intelligence is converted into multilingual audio briefings using the edge-tts neural voice synthesis engine, supporting both English and Hindi voice outputs. To avoid cloud timeout failures and synchronous processing bottlenecks, the platform implements an asynchronous polling architecture using UUID-based job orchestration. The frontend submits a /briefing request and continuously polls a /briefing-status/{job_id} endpoint until the AI-generated strategy and MP3 briefing become available. This architecture enables the system to safely execute computationally expensive large language model inference and neural voice generation workflows even on limited-resource cloud infrastructure. The completed system demonstrates the practical integration of distributed AI infrastructure, cryptographic verification, asynchronous backend engineering, financial forecasting, and multimodal intelligence synthesis within a real-world production environment. Prophet AI v1.1 represents a transition from passive financial recordkeeping to proactive survival-oriented financial intelligence. The project establishes a scalable blueprint for next-generation AI-powered fintech systems capable of delivering real-time strategic decision support for the rapidly growing global freelance economy.Keywords-Freelance finance; cashflow forecasting; stochastic simulation
Smart contract vulnerabilities in Decentralized Finance (DeFi) protocols resulted in over 1.49 billion USD in confirmed losses in 2024 alone, across 192 incidents [1]. As LLM-based vulnerability detection emerges as a promising approach to address these threats, the quality of evaluation datasets has become a critical bottleneck. Existing datasets suffer from three fundamental problems: they are built on outdated Solidity versions (e.g., v0.4) that no longer reflect modern DeFi contracts [5][6][7]; they rely on automated or LLM-generated annotations that introduce hallucination-driven label noise [9][10]; and they apply coarse single-layer labeling that fails to capture the semantic complexity of real-world business logic vulnerabilities [6][7][11][12]. We present Bastet, an expert-labeled DeFi smart contract vulnerability dataset that addresses all three problems through real-world audit findings (2021-2024), human expert annotation with discussion-based consensus, and a two-layer taxonomy of 46 Tags and 77 Subtags. Bastet comprises 4,402 findings collected from 394 Code4rena competitive audit reports spanning April 2021 to November 2024, of which 849 findings are fully annotated by white-hat security researchers from the DeFiHackLabs community. All annotations are produced through a two-annotator consensus workflow, ensuring label accuracy grounded in real-world vulnerability root causes.
Imran Hasan, Abdullah All Ahhad, Md Zamilur Rahman, Bikash Chandra Singh
Smart contracts enable decentralized applications across domains such as finance, logistics, and healthcare, but their immutable nature and complex execution logic make them highly susceptible to vulnerabilities, including reentrancy, integer overflows, and access control flaws. These weaknesses can lead to severe financial and operational losses. Traditional static or rule-based detection tools lack scalability and adaptability, while existing deep learning models often struggle with limited data, poor generalization, and the absence of actionable mitigation guidance. This paper proposes a hybrid multi-task learning framework that integrates Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks for smart contract vulnerability detection, coupled with a transformer-based Large Language Model (LLM) for root cause analysis and dynamic mitigation generation. The framework extracts spatial opcode features using CNNs and captures temporal execution patterns via LSTMs, supported by preprocessing steps that include opcode extraction, positional encoding, static and dynamic analysis features, and data augmentation. A feature fusion module consolidates spatial and temporal information, while SHAP and LIME provide interpretability by identifying features driving model predictions. The mitigation layer employs an encoder–decoder transformer to map detected vulnerabilities to their underlying causes and generate context-aware remediation strategies. Experimental results show strong performance, achieving 93% accuracy, 90% precision, and an AUC-ROC of up to 90% across multiple vulnerability categories. Beyond accurate detection, the framework delivers explainable root cause insights and tailored mitigations, offering a scalable and adaptive solution for enhancing smart contract security in modern blockchain ecosystems.
Transaction monitoring and efficient audit management have become increasingly importantin modern financial systems. Traditional centralized databases and auditing methods often face challengesrelated to security vulnerabilities, fraudulent activities, and data manipulation. A blockchain-based financialtransaction monitoring system integrates smart contracts, decentralized ledgers, and audit trails to automatefinancial operations, enhance transparency, and reduce fraud risks. The proposed architecture is implementedon Ethereum and Hyperledger Fabric platforms, enabling automated transaction validation and executionthrough smart contracts. All transactions are stored in an immutable decentralized ledger, while audit trailsare generated and maintained automatically. Simulation results demonstrate a 40–60% reduction in fraudulentactivities and up to a 70% decrease in audit processing time compared with conventional approaches. Theapplication of cryptographic algorithms and Zero-Knowledge Proofs further strengthens data security andprivacy protection. The proposed solution contributes to the improvement of financial control and auditingsystems within the framework of the digital economy.
Pedro Henrique F. S. Oliveira, Heder S. Bernardino, Saulo Moraes Villela, Edelberto Franco Silva · 6 authors
O Ethereum é uma plataforma de criptomoedas que permite a execução de contratos inteligentes, programas autônomos que operam em uma rede descentralizada. As vulnerabilidades nesses contratos representam grandes riscos financeiros e de segurança nos ecossistemas blockchain, motivando a automatização do processo de detectá-las. Este trabalho estuda a detecção de vulnerabilidades em contratos inteligentes Ethereum usando embeddings derivados de bytecode. Embeddings são representações vetoriais geradas por modelos de linguagem, que capturam as características estruturais de texto. Essas representações foram usadas como entrada para os algoritmos de regressão logística, árvore de decisão e floresta aleatória, com o fim de detectar quais contratos possuem vulnerabilidades. Os resultados mostram que os embeddings contêm informações úteis para distinguir contratos vulneráveis de não vulneráveis. O estudo também constata que a alteração da distribuição original dos dados durante o treinamento afeta significativamente o desempenho, destacando a sensibilidade das abordagens baseadas em embeddings às estratégias de amostragem.
This paper examines the use of Event-Driven Architecture (EDA) patterns to improve the optimization of financial risk evaluation in a distributed cloud-based system of finance.Today's financial system is characterized by a number of difficulties in processing high-speed data feeds in a timely manner, ensuring sub-millisecond latency and high availability.This paper proposes a decoupled system utilizing distributed event brokers and stream processors to identify market anomalies and credit risks in a timely fashion.This research utilizes a risk data set of 404 unique risk scenarios, including high-frequency trading (HFT) simulation data and credit transaction data, to measure system efficiency.The system environment utilizes Apache Kafka for event streaming, Kubernetes for cloud orchestration, and Prometheus for monitoring.The results show that event-driven architecture can improve system efficiency by eliminating traditional requestresponse processing bottlenecks.Furthermore, by utilizing distributed ledgers and serverless architecture, financial organizations can improve their risk profile granularity.The results show that by utilizing reactive programming, financial organizations can improve their risk management approach by shifting their traditional reactive approach to a proactive approach.
Central bank digital currencies (CBDCs) integrated with decentralized finance (DeFi) represent a transformative development in digital financial systems. However, there is a lack of systematic frameworks for prioritizing the determinants of effectiveness and sustainability in DeFi-integrated CBDC platform investments. This study develops an integrated multicriteria decision-making framework to identify critical evaluation criteria and rank alternative platform architectures under uncertainty. The proposed model combines objective expert weighting, interaction-sensitive criteria evaluation, and fuzzy-based alternative ranking within a unified analytical structure. The results indicate that technological infrastructure (0.168) and liquidity (0.167) are the most influential criteria, while hybrid and privacy-focused platforms emerge as the most suitable investment alternatives. These findings highlight the importance of balancing technological robustness, liquidity depth, and privacy considerations in CBDC design. The study contributes by offering a structured and uncertainty-sensitive decision framework to support strategic platform selection and policy formulation in evolving digital currency ecosystems.
Blockchain-based financial systems process billions in transactions but remain vulnerable to sophisticated fraud schemes. Current detection approaches analyze completed transactions, preventing neither fund loss nor protocol exploitation. We address this through an oracle-mediated prevention system integrating machine learning inference with smart contract execution. Training ensemble models on 12,847 Ethereum transactions with engineered features capturing gas anomalies and temporal patterns, we achieve 94.2\% fraud classification accuracy. Testnet deployment demonstrates 1.09-second response latency with 6.8\% computational overhead, contrasting favorably against prior on-chain implementations requiring 34\% overhead. Our working prototype validates practical viability for production environments where security requirements justify marginal transaction costs.
To address the challenges of topological obscurity and extreme label sparsity in large-scale Ethereum transaction networks, a novel self-supervised phishing detection framework named Eth-GBAV is proposed, integrating graph attention, broad learning, and adversarial variational inference. The framework initiates with a biased random walk strategy guided by transaction intensity and temporal dynamics to capture the initial behavioral semantics of nodes. To distill discriminative features from noisy backgrounds, a “Generative-Attention” encoding architecture is constructed, where a graph attention network aggregates weighted structural neighborhoods and a Variational Autoencoder (VAE) characterizes the underlying probability distribution of legitimate transaction patterns. By maximizing the evidence lower bound, anomalous accounts are effectively isolated through reconstruction residuals. Furthermore, the broad learning system is introduced as an efficient analytical decision layer. By mapping VAE-derived latent embeddings and reconstruction errors into an expanded high-dimensional feature space, the framework captures intricate behavioral correlations via mapping and enhancement neurons. Extensive experimental verification on two large-scale datasets demonstrates the superior performance of Eth-GBAV. On the XBlock dataset, it achieves a leading F1-score of 0.9847 and a recall of 0.9839, outperforming the most competitive state-of-the-art model by significant margins. On the Kaggle dataset, the framework maintains high robustness with an accuracy of 0.9592 and an F1-score of 0.9069.
Rapid urbanization and the exponential growth of vehicles have led to severe traffic congestion, increased travel time, fuel consumption, and environmental pollution in metropolitan cities.Traditional traffic control systems, which rely on fixed-time signals and manual monitoring, are inadequate to handle dynamic and unpredictable traffic conditions.This project proposes a Smart Traffic Management System designed to optimize traffic flow and reduce congestion using advanced technologies such as Artificial Intelligence (AI), Internet of Things (IoT), and real-time data analytics.The system integrates smart sensors, cameras, and GPS-enabled devices to continuously monitor traffic density, vehicle movement, and road conditions.Data collected from these sources is processed using machine learning algorithms to predict traffic patterns and dynamically adjust traffic signal timings.Additionally, the system provides real-time route guidance to drivers through mobile applications and digital signboards, helping to distribute traffic evenly across the road network.Emergency vehicle prioritization and incident detection mechanisms are also incorporated to enhance response efficiency and safety.
Predictive analysis has become an essential component in modern financial research and practice, driven by the rapid advancement of data analytics, machine learning, and artificial intelligence. This study aims to systematically map the intellectual structure, research trends, and key contributions in the field of predictive analysis in finance through a bibliometric approach. Data were collected from the Scopus database covering publications from 2000 to 2026 and analyzed using VOSviewer to examine co-authorship networks, citation patterns, and keyword co-occurrence. The results reveal a significant growth in research output, particularly in recent years, reflecting the increasing importance of data-driven decision-making in finance. Co-authorship analysis indicates the presence of collaborative research clusters, although the field remains partially fragmented. Citation analysis highlights that the most influential studies are those integrating advanced computational methods with practical financial applications, such as credit scoring, bankruptcy prediction, and stock market forecasting. Furthermore, keyword analysis demonstrates a clear shift from traditional statistical techniques toward machine learning, artificial intelligence, and emerging technologies such as blockchain and decentralized finance. This study contributes by providing a comprehensive overview of the evolution and current state of predictive analysis in finance, identifying key research themes and gaps. The findings suggest that future research should focus on enhancing model interpretability, integrating sustainability considerations, and expanding applications in real-time financial decision-making. Overall, this study serves as a valuable reference for researchers and practitioners seeking to understand the trajectory and future direction of predictive analytics in the financial domain.
This study rigorously formulates the complex fund-scheduling problem as a Markov decision process (MDP). It constructs a state space that integrates real-time and forecast information, an atomic action space that conforms to business logic, and a reward function that balances long-term returns against immediate risk. To address the curse of dimensionality and the credit-assignment problem in coordinated scheduling among multiple fund units, a multi-agent deep deterministic policy gradient (MADDPG) algorithm is adopted. Under a centralized-training and decentralized-execution framework, the algorithm reconciles global optimization with decentralized decision-making. In addition, a difference-reward mechanism and Kalman filtering are used to accurately measure each agent’s individual contribution and reduce the impact of environmental noise on reward signals. The results show that, compared with a static rule engine and a conventional linear programming method, the proposed deep reinforcement learning strategy reduces average daily funding costs by 50.4%, lowers the payment failure rate to 0.002%, and maintains a high liquidity buffer adequacy ratio. The strategy also demonstrates clear advantages in decision timeliness, collaborative handling of complex instructions, and self-adaptation potential, thereby providing an innovative pathway for finance-company fund scheduling to progress from intelligentization to automation.
Financial fraud across cross-border payment systems and blockchain-based transaction networks has grown in scale, sophistication, and velocity, driven by increased digitization, regulatory fragmentation, and the pseudonymous nature of decentralized infrastructures.This study presents a comprehensive examination of AI-driven anomaly detection techniques designed to address these evolving threats.From a broad perspective, the paper reviews the global financial ecosystem, highlighting vulnerabilities in traditional correspondent banking frameworks and emerging decentralized finance (DeFi) architectures.It then narrows to advanced machine learning and deep learning approaches, including supervised, unsupervised, and hybrid models such as autoencoders, graph neural networks, and reinforcement learning systems for real-time fraud detection.Particular emphasis is placed on transaction pattern analysis, behavioral profiling, and network topology modeling to uncover hidden relationships and detect anomalous activities across distributed ledgers and cross-border payment rails.The study further evaluates challenges such as data sparsity, class imbalance, adversarial manipulation, privacy constraints, and regulatory compliance, including AML and KYC requirements.By integrating AI with blockchain analytics and financial monitoring systems, the paper demonstrates how adaptive, scalable, and explainable detection frameworks can significantly enhance fraud prevention capabilities.The findings provide strategic insights for financial institutions, regulators, and fintech developers aiming to strengthen global financial security.
Decentralized Finance (DeFi) lending protocols like Aave v3 rely on over-collateralization to secure loans, yet users frequently face liquidation due to volatile market conditions. Existing risk management tools utilize static health-factor thresholds, which are reactive and fail to distinguish between administrative "dust" cleanup and genuine insolvency. In this work, we propose an autonomous agent that leverages time-to-event (survival) analysis and moves beyond prediction to execution. Unlike passive risk signals, this agent perceives risk, simulates counterfactual futures, and executes protocol-faithful interventions to proactively prevent liquidations. We introduce a return period metric derived from a numerically stable XGBoost Cox proportional hazards model to normalize risk across transaction types, coupled with a volatility-adjusted trend score to filter transient market noise. To select optimal interventions, we implement a counterfactual optimization loop that simulates potential user actions to find the minimum capital required to mitigate risk. We validate our approach using a high-fidelity, protocol-faithful Aave v3 simulator on a cohort of 4,882 high-risk user profiles. The results demonstrate the agent's ability to prevent liquidations in imminent-risk scenarios where static rules fail, effectively "saving the unsavable" while maintaining a zero worsening rate, providing a critical safety guarantee often missing in autonomous financial agents. Furthermore, the system successfully differentiates between actionable financial risks and negligible dust events, optimizing capital efficiency where static rules fail.
As cross-chain interoperability advances, decentralized finance (DeFi) protocols enable illicit funds to be reorganized into uniform liquid assets that flow throughout the cryptocurrency market. Such operations can bypass monitoring targeted at individual blockchains and thereby weaken current regulatory frameworks. Motivated by these, we introduce UniDetect, a multi-chain cryptocurrency fraud account detection method based on large language models (LLMs). Specifically, we use domain knowledge to guide the LLM to generate general transaction summary texts applicable to heterogeneous blockchain accounts, which serve as evidence for fraud account detection. Furthermore, we introduce a two-stage alternating training strategy to continuously and dynamically enhance the multimodal joint reasoning for detecting fraudulent accounts based on both the textual evidence and the transaction graph patterns. Experiments on multiple blockchains show that UniDetect outperforms existing methods 5.57% to 7.58% in Kolmogorov-Smirnov (KS). For cross-chain zero-shot detection, UniDetect identifies over 94.58% of fraudulent accounts. It also generalizes well to non-blockchain data, delivering a 6.06% improvement in F1 over existing methods. The dataset and source code are available at https://github.com/msy0513/UniDetect.
Fraudulent activities on blockchain networks threaten the integrity and reliability of decentralized finance ecosystems. Accurately identifying malicious nodes such as phishing or ransomware addresses, within large-scale blockchain transaction graphs remains a critical challenge due to their dynamic, sparse, and continuously evolving topologies. Transfer learning offers a powerful paradigm for fraud detection because many fraudulent schemes, including ransomware and phishing, are often orchestrated by overlapping actor groups that share behavioral and structural patterns across networks. Leveraging these shared representations enables knowledge transfer from previously observed fraud types to emerging ones. However, the complex and multi-modal nature of digital financial systems introduces substantial challenges for graph-based transfer learning. Fraudulent activities are shaped by diverse modalities including graph structure, transaction sequences, temporal price dynamics, and textual metadata, while distributional shifts frequently occur across time and platforms. Existing graph transfer learning methods struggle to model such multi-modal dependencies and to align divergent feature distributions. To tackle these challenges, we develop a Multi-mOdal Enhanced Graph Transfer Learning (MOE-GTL) framework which incorporates graph, temporal, and textual modalities for fraudulent node detection. We further introduce Temporal-aware Maximum Mean Discrepancy (TMMD), a regularization mechanism that explicitly aligns multi-modal feature distributions between source and target graphs over time. Extensive experiments reveal that our MOE-GTL model notably improves the accuracy of fraudulent node classifications on Ethereum and Solana transaction graphs.
Against the complex characteristics of the Ethereum transaction network and the limitations of existing graph embedding methods based on random walks, which fail to effectively capture transaction temporal dynamics and the flow of funds, we propose a fraud detection algorithm for Ethereum, ETX2Vec (Ethereum Transactions (TX) to Vector), which improves upon transaction subgraph construction and random walk strategies. First, in terms of transaction subgraph construction, we extract the first-order predecessor and successor neighboring nodes of the target node to reconstruct the transaction subgraph, enabling the random walk to effectively capture the complete flow of funds. Second, in the design of the random walk strategy, we introduce two key improvements: (1) the next node is selected based on the non-decreasing principle of transaction timestamps, effectively capturing the temporal dynamics of transactions within the network, and (2) a biased random walk strategy is designed based on both transaction timestamps and amounts, with a parameter α introduced to control the weighting of these factors when calculating transition probabilities. Experimental results show that ETX2Vec achieves an average performance of 96.04% in downstream node classification tasks, outperforming the best model in similar studies by 3.74%, and even surpassing neural network models such as GAT and GCN. This demonstrates that ETX2Vec is more effective at understanding and processing the Ethereum transaction network, leading to the learning of high-quality node embedding vectors.
Traditional consensus mechanisms, such as Proof of Stake (PoS), increasingly reveal an excessive dependency on large liquidity providers. Although the Proof of Liquidity (PoL) mechanism serves as a critical paradigm for incentivizing sustained liquidity provision and ensuring market stability, its transition from asset staking to active liquidity management significantly increases the complexity of underlying smart contract economic models and interaction logic. This renders hidden liquidity logic flaws difficult to detect via traditional methods, seriously threatening the system stability and user asset security of mainstream DeFi and emerging PoL ecosystems. To address this, we propose the LiquiLM framework, which integrates Large Language Models (LLMs) with a Dynamic Co-Attention Network (DCN). By establishing a dynamic interaction between liquidity-critical contracts and flaw descriptions, the framework effectively bridges the semantic gap between underlying code implementations and high-level liquidity intents. We evaluate the performance of LiquiLM on 1,490 validation contracts (covering precision, recall, specificity, and F1-score). The results show that it achieves significant effectiveness in auditing and explaining liquidity flaws: in experiments using Gemini 3 Pro and GPT-4o as backbone models, respectively, the F1-scores both exceed 90%. Furthermore, through an in-depth audit of 1,380 real-world PoL and Ethereum economic contracts, LiquiLM successfully identifies 238 high-risk contracts and assists in discovering 10 vulnerabilities that have received CVE certification.
O presente artigo examina as distinções entre finalidade probabilística e finalidade absoluta em sistemas blockchain, bem como suas implicações para o desenho e a operação de aplicações financeiras que visam a replicar ou substituir infraestruturas tradicionais de liquidação. Em cadeias que operam sob finalidade probabilística – modelo historicamente associado a protocolos baseados em Prova de Trabalho (Proof-of-Work) – o grau de irreversibilidade de uma transação cresce à medida que novos blocos são adicionados sobre o bloco que a contém, de modo que a probabilidade de reversão tende assintoticamente a zero sem, contudo, alcançar garantia determinística, o que justifica a prática de mercado de aguardar múltiplas confirmações antes de considerar a liquidação efetivamente concluída. Em contrapartida, cadeias dotadas de finalidade absoluta – também denominada finalidade instantânea – usualmente implementadas sobre protocolos de tolerância a falhas bizantinas (BFT) ou em arquiteturas híbridas que combinam Prova de Participação (PoS) e BFT, oferecem irreversibilidade assim que um bloco é atestado por um superconjunto qualificado de validadores, aproximando-se das expectativas de definitividade inerentes a sistemas de liquidação financeira tradicionais. A metodologia adotada combina revisão conceitual das diferentes acepções de finality em mecanismos de consenso, análise de documentação técnica de protocolos BFT – a exemplo de Tendermint, IBFT e QBFT – e discussão de relatórios recentes sobre risco de liquidação e finality aplicáveis à tokenização de ativos do mundo real (Real World Assets – RWA) em infraestruturas on-chain. Os resultados obtidos sinalizam que, embora a finalidade probabilística se mostre adequada a pagamentos de varejo e transferências de valor moderado, aplicações financeiras de maior montante, processos de tokenização de ativos e infraestruturas de mercado requerem, na prática, garantias mais robustas de irreversibilidade, com frequência combinando finalidade técnica e mecanismos jurídicos de mitigação de risco de liquidação. Conclui-se que a opção entre os dois modelos de finalidade encerra trade-offs relevantes entre segurança, velocidade de confirmação, complexidade de protocolo e conformidade regulatória, e que o desenho de aplicações financeiras em ambiente Web3 deve considerar explicitamente essas diferenças ao definir janelas de liquidação, políticas de gerenciamento de risco e estratégias de integração com o sistema financeiro tradicional.
Abstract To address issues in traditional patent valuation—such as subjective selection of dimensional metrics, weak sensitivity to high-dimensional transaction data noise, and insufficient correlation between evaluation indicators and dimensions—this study proposes a smart contract-based patent value assessment model. Firstly, existing patent valuation theories and techniques undergo systematic deconstruction and multidimensional efficacy assessment. Leveraging big data technology, a four-dimensional optimal framework integrating "technology-market-legal-risk" dimensions is constructed. Secondly, an enhanced non-negative matrix factorization algorithm (S-NMF) is designed. By incorporating diagonal matrices and fused regularization parameters, this algorithm maps the four-dimensional optimal framework into 14 quantifiable metrics using Hyperledger Fabric consortium blockchain transaction data. This addresses the core limitation of classical NMF algorithms—the inability to adjust dimension weights—enabling flexible weighting control to meet differentiated valuation needs across diverse patent application scenarios. Finally, performance analysis and simulation experiments were conducted on the patent value assessment model, comparing it with the traditional NMF algorithm. Results demonstrate that this model outperforms traditional models in both noise robustness and dimensional correlation, effectively supporting patent value assessment needs across multiple scenarios.
This study examines the development and intellectual structure of fraud detection research through a bibliometric analysis. Using data extracted from a major scientific database and analyzed with bibliometric visualization tools, the study maps publication trends, influential contributors, and thematic evolution within the field. The findings reveal that fraud detection research is strongly centered on machine learning and increasingly shaped by advances in deep learning, neural networks, and data-driven approaches. At the same time, the field has expanded beyond traditional financial contexts into broader digital ecosystems, including cybersecurity, blockchain, and data privacy. The analysis also highlights a clear shift from conventional statistical methods toward more adaptive and complex models capable of handling large-scale and interconnected data. In addition, emerging themes such as predictive analytics, risk management, and decentralized finance indicate a growing orientation toward real-world application and decision-making. Overall, the study provides a comprehensive overview of the research landscape, identifies key trends and gaps, and offers directions for future research, particularly in integrating technological innovation with practical, ethical, and system-level considerations.