
In 2015, Lemos, Daniel, and Benatallah introduced a comprehensive taxonomy of Web service composition approaches, techniques, and tools. Given that ten years have passed and the recent emergence of artificial intelligence techniques and Large Language Models (LLMs) for service composition and systems engineering, there is a need to update such a classification. To this end, we perform a Systematic Literature Review (SLR) on natural language-based automated service composition approaches and filter for approaches not covered within the existing taxonomy. Specifically, we study 234 papers published from 2014 to May 2026 and identify 20 novel approaches that the original taxonomy inadequately covers, given the technological shifts. The result of the present study is a revised taxonomy integrating eight additional subcategories. The revised taxonomy provides a structured framework, making it suitable for future work on automated service composition with LLMs.
Financial fraud has become increasingly covert, well-organized, and cross-domain, posing significant challenges for traditional methods that struggle to capture the complex heterogeneous relationships and node influence within financial networks. Consequently, a persistent trade-off exists between precision and recall. To address this problem, this paper proposes a Propagation Influence Quantification (PIQ) enhanced Heterogeneous Graph Transformer (HGT) model, named PIQ-HGT, together with an integrated ‘detection-explanation-decision’ support framework. Specifically, we construct a heterogeneous information network from financial transaction data, introduce a propagation influence quantification feature that captures propagation capability, and fuse it with original attributes to enhance node representation. Then, a fused-feature-driven HGT model with residual connections is designed to model complex heterogeneous relationships. Furthermore, a multi-dimensional explainability framework is built from feature importance, community risk, graph structure, and node levels to provide evidence-based risk explanations. Experiments on three public datasets -DGraphFin, IBM AML, and Elliptic++—demonstrate that PIQ-HGT consistently outperforms baseline models (GCN, GAT, GraphSAGE, HAN, RGCN, FastGTN, SCENE and standard HGT) in terms of precision, recall, F1-score, and AUC. In particular, PIQ-HGT achieves superior detection of high-risk nodes and communities, effectively traces fraud propagation paths, and provides quantitative and visual decision support for financial anti-fraud.
With the rapid advancement of emerging Internet technologies, the number and diversity of land cover services have increased substantially, leading to the development of an open land cover service ecosystem. Consequently, the automated orchestration of fine-grained services to construct land cover service chains has become an important research topic. However, Web-based land cover service composition environments are often subject to unpredictable anomalies, which may result in service execution failures or access interruptions. Therefore, an effective anomaly handling mechanism is essential to ensure the reliability and robustness of service composition. To address this challenge, this paper proposes an adaptive land cover service composition mechanism based on a service network. The proposed approach integrates land cover characteristics with service network properties to develop an adaptive model, an anomaly classification and decision-making framework, and an adaptive service composition algorithm. Furthermore, a Web-based land cover service composition system is implemented and evaluated through a land cover change detection case study. Experimental results demonstrate that the proposed method improves the success rate of service execution and enhances the adaptability and reliability of land cover service composition.
Unstructured financial texts provide event-related signals for event-aware forecasting in intelligent financial services, but transforming heterogeneous texts into structured predictive representations remains challenging. Existing studies often insufficiently distinguish event influence scopes, inadequately model event interactions, and rely on shallow fusion between textual events and numerical market signals, thereby limiting their use in service-oriented financial analytics. To address these limitations, we propose an LLM-based event-driven service framework for intelligent financial forecasting. The framework transforms financial news and corporate announcements into structured event representations and integrates them with numerical market features. A schema-guided extraction pipeline identifies predefined financial event types and structured arguments, organizes events into local and global categories, and encodes them using template-based descriptions. A multi-scale convolutional sparse attention module captures event dependencies and cross-level interactions, whereas adaptive gated fusion integrates event and market representations. Stock movement forecasting is used as a representative downstream service task. Experiments on CSI 300 constituent stocks from 2022 to 2024 show that the proposed method outperforms representative baselines, achieving an ACC of 56.06
Large language models now serve enterprise systems, where they plan work, execute handoffs, review artifacts, and improve prior outputs rather than only answer isolated chat requests. Most safety benchmarks test single prompts or jailbreak dialogues and miss how unsafe intent can flow through workflow history, inherited subtasks, and intermediate artifacts. We present JailFlowEval, a workflow-safety benchmark that turns harmful intents into stage-specific records for four common workflow contexts: plan, exec, review, and improve. Its two-phase pipeline builds workflow state and rewrites refusal-triggering text while keeping the unsafe goal; the released high-quality subset contains 744 workflow evaluation records. Tests on 10 models show that workflow-aware subtasks reveal unsafe continuation risks missed by prompt-level evaluation. The main risk context varies by model and metric. This setup maps risk by stage and helps place safeguards where unsafe goals persist as work moves through later service steps after the initial user request is processed.