Real bank-transaction data are rarely available to non-regulatory parties due to privacy and compliance constraints, while public substitutes often mismatch financial semantics, limiting the transferability and reproducibility of anomaly-detection research in financial settings. To address this, we release F2-Gen: an open-source, scenario-driven, configurable web platform that generates synthetic data covering both tabular transaction records and transaction graphs for modeling and evaluating financial-fraud behaviors. F2-Gen provides parameterized fraud scenarios, interactive visual analysis, automated screening to remove label-revealing fields, tutorials, and batch data export. Use cases include anomaly detection, graph learning, and teaching demonstrations; standard benchmarking is supported via a screened feature subset, ready-to-run baseline scripts, and sanity/controllability checks. Project page: https://sethgu.github.io/FinancialFraudDataGenerator/. Code: https://github.com/sethGu/FinancialFraudDataGenerator.

Unsupervised graph anomaly detection seeks to discover rare patterns in graph-structured data, enabling applications across diverse fields such as financial transactions, e-commerce, and beyond. Although recent methods based on graph autoencoder and graph contrastive consistency show considerable promise in this field, the former primarily addresses global anomalies, whereas the latter is more focused on local anomalies. This distinction limits their capacity to fully capture the abnormalities of nodes in the graph. Moreover, challenges such as overfitting of abnormal patterns and the introduction of unknown noise from random walk-based subgraph sampling remain prevalent. These issues contribute to suboptimal model performance. To tackle these challenges, we propose Local and Global Contrastive Networks (LGCN). LGCN first extracts graph features using a feature encoder module and subsequently projects them into local and global feature subspaces. The global information learning module mitigates overfitting in reconstruction techniques to abnormal patterns by simultaneously accounting for both the similarities and differences between nodes. The local information learning module extracts anomaly information by evaluating the consistency between the center node and its corresponding context features. Extensive experiments across eight datasets demonstrate that LGCN outperforms existing methods in unsupervised graph anomaly detection.

Knowledge Base Question Answering (KBQA) aims to answer natural language questions by reasoning over large-scale structured Knowledge Bases (KBs). Among existing approaches, semantic parsing-based methods have emerged as a mainstream solution, where Large Language Models (LLMs) are employed to translate questions into structured graph queries such as Logical Forms (LFs). However, this paradigm faces two critical challenges: (1) The complex semantic mapping and graph retrieval operations render direct one-shot LF generation difficult; (2) LLMs suffer from inherent hallucination issues, generating semantically plausible-seeming but factually incorrect or invalid LFs, which are non-executable. To address these challenges, this article proposes HP-Corr , a novel framework that integrates H op-wise P lanning with iterative explainable self- Corr ection for faithful knowledge reasoning. Specifically, the system utilizes a fine-tuned open source LLM for query planning and explainable self-correction. The query planner generates reasoning paths hop-by-hop, while an explainable self-correction provides hop-wise feedback, enabling interpretable path editing based on existing reasoning paths and retrieved KB knowledge. By introducing the dual-module cooperative architecture, our system performs iterative plan-then-correct to refine query paths progressively, ensuring answer reliability and LFs executability. Experimental results demonstrate significant improvements, with our approach achieving higher accuracy while substantially reducing the search space, particularly in complex multi-hop KBQA scenarios.

Although the fine-tuning-based deep transfer learning method performs well in adapting a pre-trained model to downstream tasks, it struggles in non-stationary environments where the data distribution changes dynamically over time. In such scenarios, the trained model rapidly becomes obsolete. Directly fine-tuning it for new environments will inevitably lead to negative transfer, degrading the model's generalization performance. Inspired by neuroscience, some existing studies have proposed a so-called “unlearning-and-relearning” training paradigm to alleviate this issue. However, existing methods face limitations; for example, they may destroy general features in shallow layers and lack adaptability when performing deep-layer unlearning. This paper proposes a Layer-wise Unlearning (LwU) method. Our approach first identifies the layers requiring unlearning by estimating the transferability of each layer in the trained model. Then, within the selected layers, it preserves high-sensitivity parameters and re-initializes low-sensitivity ones based on a parameter sensitivity analysis to achieve unlearning. Extensive experiments on synthetic, real-world, and image datasets demonstrate that LwU effectively overcomes negative transfer and substantially improves the model's generalization performance in new environments. The code of the proposed method and data are available at https://github.com/mlmmwym/LwU.

There are a large number of text processing tasks in web applications, such as sentiment classification, summary extraction, and question answering. Recently, fine-tuning pre-trained language models (PLMs) to adapt to downstream text-processing tasks has attracted much attention. However, due to the differences in data, model, and tasks between the pre-training and fine-tuning processes, the fine-tuning process may suffer from catastrophic forgetting of pre-training knowledge, which may implicitly limit the model's performance and generalization ability. To address these challenges, we propose a novel dual-model framework, termed as consistency alignment (CoAi). The insight of CoAi lies in building an auxiliary model that simulates the distribution of pre-training knowledge in real-time according to the current task, and co-training the task-specific model and the auxiliary model to balance the pre-training knowledge and task-specific knowledge during fine-tuning. Specifically, the auxiliary model is constructed on-the-fly to maintain the pre-training knowledge. Subsequently, CoAi simulates the pre-training process by performing distributional exploration in the parameter space, which is built upon our novel insight into the transformation between data and model parameter space. However, the objectives leveraged to construct the auxiliary model lead to the misalignment between the pre-training and task-specific knowledge. To alleviate the inconsistency, we employ an auxiliary variable to align the prediction distribution of the task-specific and the auxiliary models, inspired by constrastive clustering. We validate the effectiveness of CoAi on nine classic classification tasks and three generation tasks, showing consistent and significant improvements compared with state-of-the-art methods.

Graph anomaly detection plays a crucial role in identifying nodes that deviate significantly from normal patterns within a graph, with applications spanning various domains such as detection of authorship fraud and rumor propagation. Traditional methods primarily focus on aggregating information from neighboring nodes and reconstructing the central node based on these aggregated features. The anomaly degree is then calculated by comparing the reconstructed features with the original ones. Despite their effectiveness, these methods face limitations due to the constraints of device performance and the need to protect user privacy. In reality, graph data is often partitioned and distributed across different local clients, which leads to isolated client subgraphs. This partitioning results in incomplete feature aggregation, as the connections between subgraphs are missing, ultimately reducing the performance of anomaly detection models. To overcome these challenges, a federated graph anomaly detection approach based on disentangled representation learning is proposed. This method separates node features into two distinct components: intrinsic features and subgraph style features. By identifying outliers within the subgraph style features, a set of pseudo-nodes is generated and shared across the entire graph. These pseudo-nodes simulate connections between otherwise isolated subgraphs, which enables more comprehensive aggregation of intrinsic features from neighboring nodes. In addition, conditional variational autoencoders (CVAE) are employed alongside contrastive learning strategies to alleviate class imbalance and achieve effective feature disentanglement. These techniques help ensure that anomalous nodes are detected more accurately despite the inherent challenges of federated graph systems. Extensive experiments conducted on six diverse datasets provide compelling evidence of the proposed method's superior performance in federated graph anomaly detection, highlighting its ability to effectively handle incomplete graph structures while maintaining data privacy.

Concept drift is a phenomenon where the distribution of data streams changes over time. When this happens, model predictions become less accurate. Hence, models built in the past need to be re-learned for the current data. Two design questions need to be addressed in designing a strategy to re-learn models: which type of concept drift has occurred, and how to utilize the drift type to improve re-learning performance. Existing drift detection methods are often good at determining when drift has occurred. However, few retrieve information about how the drift came to be present in the stream. Hence, determining the impact of the type of drift on adaptation is difficult. Filling this gap, we designed a framework based on a lazy strategy called Type-Driven Lazy Drift Adaptor (Type-LDA). Type-LDA first retrieves information about both how and when a drift has occurred, then it uses this information to re-learn the new model. To identify the type of drift, a drift type identifier is pre-trained on synthetic data of known drift types. Furthermore, a drift point locator locates the optimal point of drift via a sharing loss. Hence, Type-LDA can select the optimal point, according to the drift type, to re-learn the new model. Experiments validate Type-LDA on both synthetic data and real-world data, and the results show that accurately identifying drift type can improve adaptation accuracy.

Concept drift is a phenomenon that the distribution of data streams changes with time. When this happens, model predictions become less accurate. Hence, concept drift needs to be detected and adapted. Existing drift detection methods are good at determining when drift has occurred, but few retrieve information about how the drift came to be present in the stream, i.e., what type of drift has occurred. Hence, discussing the impact of the type of drift on adaptation is a difficult thing. To fill this gap, we propose a pre-trained framework for training a drift detector called a type-driven lite concept drift detector (Type-LDD) that retrieves information about both when and how a drift has occurred. In our proposed pre-trained framework, the Type-LDD including a drift-type identifier and a drift-point locator was based on a synthetic dataset containing a range of drift types. When repurposing the pre-trained model for detecting new data streams, a knowledge distillation module fine-tunes the proposed Type-LDD to speed up inference and keep detection accuracy. The proposed Type-LDD is validated on both synthetic data and real-world data, and demonstrated that accurately identifying the type of drift that has occurred can improve adaptation accuracy.

Since support vector regression (SVR) is a flexible regression algorithm, its computational complexity does not depend on the dimensionality of the input space, and it has excellent generalization capability. However, a central assumption with SVRs is that all the required data is available at the time of construction, which means these algorithms cannot be used with data streams. Incremental SVR has been offered as a potential solution, but its accuracy suffers with noise and learning speeds are slow. To overcome these two limitations, we propose a novel incremental regression algorithm, called online robust support vector regression (ORSVR). ORSVR solves nonparallel bound functions simultaneously. Hence, the large quadratic programming problem (QPP) in classical v-SVR are decomposed into two smaller QPPs. An incremental learning algorithm then solves each QPP step-by-step. The results of a series of comparative experiments demonstrate that the ORSVR algorithm efficiently solves regression problems in data streams, with or without noise, and speeds up the learning process.