2025· International Journal of Machine Learning and Predictive Analytics· 0 citations
TL;DR
This paper reviews hybrid KG–LLM frameworks for predictive analytics, highlighting graph embeddings, Retrieval-Augmented Generation (RAG), transformer-based reasoning, and contextual embedding fusion to improve prediction accuracy, interpretability, and robustness.
Abstract
Artificial Intelligence has significantly advanced predictive analytics, but traditional machine learning and deep learning models often struggle to integrate structured knowledge and provide explainable reasoning. Hybrid Knowledge Graph–Large Language Model (KG–LLM) architectures address these limitations by combining the structured semantic representation of Knowledge Graphs with the contextual reasoning capabilities of LLMs. This paper reviews hybrid KG–LLM frameworks for predictive analytics, highlighting graph embeddings, Retrieval-Augmented Generation (RAG), transformer-based reasoning, and contextual embedding fusion to improve prediction accuracy, interpretability, and robustness. The framework supports applications including healthcare, finance, cybersecurity, manufacturing, and customer analytics. Performance is evaluated using standard metrics such as Accuracy, Precision, Recall, F1-Score, AUC, and MAE, demonstrating superior results over standalone approaches. The paper also discusses scalability, computational challenges, explainable AI, federated learning, multimodal knowledge graphs, and future directions for trustworthy AI-driven predictive analytics.
Artificial Intelligence (AI) has significantly advanced predictive analytics across domains such as healthcare, finance, manufacturing, cybersecurity, and smart cities. While machine learning and deep learning models achieve strong predictive performance, they often lack structured knowledge integration and semantic reasoning. Knowledge Graphs (KGs) provide structured representations of entities and relationships but face challenges such as incomplete knowledge and limited adaptability. Conversely, Large Language Models (LLMs) offer powerful language understanding and contextual reasoning but may generate hallucinations and lack transparent reasoning. Hybrid Knowledge Graph–Large Language Model (KG–LLM) architectures address these limitations by combining symbolic reasoning with neural intelligence. This paper presents a comprehensive framework integrating graph embeddings, retrieval-augmented generation (RAG), transformer-based reasoning, attention mechanisms, and contextual embedding fusion to improve prediction accuracy, explainability, and robustness. The proposed approach supports applications including disease prediction, fraud detection, financial forecasting, predictive maintenance, customer analytics, and cybersecurity. Performance is evaluated using metrics such as Accuracy, Precision, Recall, F1-Score, AUC, and MAE, demonstrating superior results compared with standalone ML, KG, and LLM models. The study also discusses challenges, scalability, computational requirements, and future directions, including multimodal knowledge graphs, federated learning, explainable AI, and autonomous knowledge reasoning for trustworthy predictive analytics.
Mahabala H.N· International Journal of Int...· 0 citations
Predictive analytics is rapidly evolving through Artificial Intelligence (AI), particularly with the emergence of foundation models that enable scalable, transferable, and context-aware intelligence across multiple domains. Unlike traditional machine learning models, foundation models leverage large-scale multimodal pretraining and efficient task-specific adaptation, enabling superior reasoning, zero-shot learning, and cross-domain knowledge transfer. This paper proposes the Foundation Model-Based Predictive Analytics Framework for Multi-Domain Decision Intelligence (FMPA-MDI), an integrated architecture that combines heterogeneous data acquisition, multimodal preprocessing, semantic representation learning, transformer-based predictive reasoning, retrieval-augmented learning, knowledge graph integration, explainable AI (XAI), and intelligent decision optimization. The framework supports structured and unstructured data while incorporating transfer learning, attention mechanisms, semantic embeddings, and continuous feedback for adaptive decision-making. Mathematical formulations model feature representation, semantic similarity, predictive confidence, and optimization. The proposed framework enhances prediction accuracy, scalability, interpretability, and computational efficiency, providing a robust foundation for next-generation intelligent decision support across healthcare, finance, manufacturing, smart cities, and other enterprise domains.
Seshagiri N· International Journal of Mac...· 0 citations
Graph Foundation Models (GFMs) enable universal representation learning from heterogeneous graph-structured data through large-scale self-supervised pretraining. Unlike traditional Graph Neural Networks (GNNs), GFMs learn transferable structural, semantic, and contextual knowledge across multiple domains, including healthcare, finance, manufacturing, cybersecurity, and smart cities, making them highly effective for cross-domain knowledge integration. This paper presents an intelligent multi-layer GFM framework for integrating heterogeneous knowledge graphs and supporting scalable cross-domain analytics. The framework combines graph representation learning, knowledge graph embedding, transformer-based graph encoders, self-supervised contrastive learning, and domain adaptation to perform semantic alignment, feature extraction, graph embedding optimization, and downstream reasoning within a unified environment. A cross-domain integration pipeline automatically aligns entities, relationships, semantics, and graph topologies from diverse data sources while transformer-based graph attention captures both local and global structural dependencies. The proposed framework is evaluated using metrics such as knowledge integration accuracy, graph embedding accuracy, node classification, link prediction, semantic consistency, computational efficiency, scalability, and inference latency. Experimental results demonstrate improved cross-domain representation learning, enhanced transfer learning, reduced feature engineering, and lower dependence on labeled data compared with conventional graph learning approaches. The framework provides a scalable foundation for graph-based artificial intelligence and has applications in biomedical knowledge discovery, financial fraud detection, industrial digital twins, recommendation systems, cybersecurity intelligence, scientific literature mining, and smart governance. Overall, Graph Foundation Models offer a promising solution for universal graph intelligence, enabling accurate cross-domain reasoning, predictive analytics, and explainable decision-making.
Seppo Linnainmaa, A. Salomaa· International Journal of Eme...· 0 citations
The rapid growth of unstructured and heterogeneous data in modern information systems has created a need for intelligent methods to extract, organize, and utilize knowledge effectively. AI-based Knowledge Graphs (KGs) address this challenge by representing entities and their relationships in a semantically rich graph structure, enabling advanced reasoning and decision support. By integrating machine learning, natural language processing, and deep learning, KGs automate entity extraction, relationship identification, and knowledge inference, improving decision-making across domains such as healthcare, finance, e-commerce, and governance. This paper presents a framework combining data preprocessing, ontology development, graph embedding, and inference techniques. Experimental results show that AI-driven knowledge graphs significantly enhance decision accuracy, reduce ambiguity, and improve interpretability, achieving up to 85–92% higher decision efficiency compared to traditional methods. Future research focuses on scalability, explainability, and integration with emerging technologies like IoT and edge computing.
Venkatesh Iyer, Nandhini Ravi· International Journal of Art...· 0 citations
Predictive intelligence enables systems to forecast future events using historical data, domain knowledge, and advanced analytics. Traditional approaches are either knowledge-driven, offering interpretability and reasoning, or data-driven, providing strong learning capabilities but facing challenges in explainability and adaptability. Hybrid predictive intelligence combines both paradigms to overcome their limitations. The proposed framework includes four stages: knowledge acquisition, data preprocessing, hybrid model integration, and predictive decision support. By integrating expert knowledge with machine learning techniques, it improves prediction accuracy, reliability, transparency, and decision-making. Applications span healthcare, industrial automation, cybersecurity, finance, smart cities, and intelligent transportation systems. Comparative studies show that hybrid models outperform conventional approaches in accuracy, robustness, and interpretability. Future developments in federated learning, digital twins, graph neural networks, and autonomous reasoning are expected to further enhance predictive intelligence for next-generation intelligent systems.
Karen Lewis, Steven Young· International Journal of App...· 0 citations
Machine Learning models are widely used for automating classification tasks by extracting statistical patterns from data. However, their performance deteriorates if the data distribution changes, making them ill-suited to handle uncertain and evolving information. Moreover, they provide limited support for integrating prior knowledge. To address these limitations, we present CLARK (Closed-loop Learning for Adaptive Reasoning over Knowledge Graphs), a framework that integrates knowledge graphs, symbolic rule mining, and probabilistic reasoning under the Logic Programs with Markov Logic Networks (LP$^{\text{MLN}}$) formalism. Starting from CACTUS-derived KGs, CLARK translates graph structure into an LP$^{\text{MLN}}$ program and iteratively enriches it with candidate rules proposed by symbolic learners. These rules are calibrated through probabilistic weight learning, enabling reasoning under uncertainty and refinement of the underlying graph structure. We evaluate CLARK on two medical datasets, analysing both rule quality and downstream classification performance. Results demonstrate that CLARK leads to improved classification performance and more generalisable inference. Overall, CLARK provides a principled approach to constructing adaptive, interpretable, knowledge-driven models for classification.
Yousef Khan, Luca Gherardini, M. Maratea et al.· 0 citations