Pour voir les autres types de publications sur ce sujet consultez le lien suivant : Knowledge Graph-based Retrieval Augmented Generation (KG-RAG).
Articles de revues sur le sujet « Knowledge Graph-based Retrieval Augmented Generation (KG-RAG) »
Créez une référence correcte selon les styles APA, MLA, Chicago, Harvard et plusieurs autres
Consultez les 50 meilleurs articles de revues pour votre recherche sur le sujet « Knowledge Graph-based Retrieval Augmented Generation (KG-RAG) ».
À côté de chaque source dans la liste de références il y a un bouton « Ajouter à la bibliographie ». Cliquez sur ce bouton, et nous générerons automatiquement la référence bibliographique pour la source choisie selon votre style de citation préféré : APA, MLA, Harvard, Vancouver, Chicago, etc.
Vous pouvez aussi télécharger le texte intégral de la publication scolaire au format pdf et consulter son résumé en ligne lorsque ces informations sont inclues dans les métadonnées.
Parcourez les articles de revues sur diverses disciplines et organisez correctement votre bibliographie.
1
Tian, Shiyu, Shuyue Xing, Xingrui Li, Yangyang Luo, Caixia Yuan, Wei Chen, Huixing Jiang et Xiaojie Wang. « A Systematic Exploration of Knowledge Graph Alignment with Large Language Models in Retrieval Augmented Generation ». Proceedings of the AAAI Conference on Artificial Intelligence 39, no 24 (11 avril 2025) : 25291–99. https://doi.org/10.1609/aaai.v39i24.34716.
Texte intégralRésumé :
Retrieval Augmented Generation (RAG) with Knowledge Graphs (KGs) is an effective way to enhance Large Language Models (LLMs). Due to the natural discrepancy between structured KGs and sequential LLMs, KGs must be linearized to text before being inputted into LLMs, leading to the problem of KG Alignment with LLMs (KGA). However, recent KG+RAG methods only consider KGA as a simple step without comprehensive and in-depth explorations, leaving three essential problems unclear: (1) What are the factors and their effects in KGA? (2) How do LLMs understand KGs? (3) How to improve KG+RAG by KGA? To fill this gap, we conduct systematic explorations on KGA, where we first define the problem of KGA and subdivide it into the graph transformation phase (graph-to-graph) and the linearization phase (graph-to-text). In the graph transformation phase, we study graph features at the node, edge, and full graph levels from low to high granularity. In the linearization phase, we study factors on formats, orders, and templates from structural to token levels. We conduct substantial experiments on 15 typical LLMs and three common datasets. Our main findings include: (1) The centrality of the KG affects the final generation; formats have the greatest impact on KGA; orders are model-dependent, without an optimal order adapting for all models; the templates with special token separators are better. (2) LLMs understand KGs by a unique mechanism, different from processing natural sentences, and separators play an important role. (3) We achieved 7.3% average performance improvements on four common LLMs on the KGQA task by combining the optimal factors to enhance KGA.
2
Hamza, Ameer, Abdullah, Yong Hyun Ahn, Sungyoung Lee et Seong Tae Kim. « LLaVA Needs More Knowledge : Retrieval Augmented Natural Language Generation with Knowledge Graph for Explaining Thoracic Pathologies ». Proceedings of the AAAI Conference on Artificial Intelligence 39, no 3 (11 avril 2025) : 3311–19. https://doi.org/10.1609/aaai.v39i3.32342.
Texte intégralRésumé :
Generating Natural Language Explanations (NLEs) for model predictions on medical images, particularly those depicting thoracic pathologies, remains a critical and challenging task. Existing methodologies often struggle due to general models' insufficient domain-specific medical knowledge and privacy concerns associated with retrieval-based augmentation techniques. To address these issues, we propose a novel Vision-Language framework augmented with a Knowledge Graph (KG)-based datastore, which enhances the model's understanding by incorporating additional domain-specific medical knowledge essential for generating accurate and informative NLEs. Our framework employs a KG-based retrieval mechanism that not only improves the precision of the generated explanations but also preserves data privacy by avoiding direct data retrieval. The KG datastore is designed as a plug-and-play module, allowing for seamless integration with various model architectures. We introduce and evaluate three distinct frameworks within this paradigm: KG-LLaVA, which integrates the pre-trained LLaVA model with KG-RAG; Med-XPT, a custom framework combining MedCLIP, a transformer-based projector, and GPT-2; and Bio-LLaVA, which adapts LLaVA by incorporating the Bio-ViT-L vision model. These frameworks are validated on the MIMIC-NLE dataset, where they achieve state-of-the-art results, underscoring the effectiveness of KG augmentation in generating high-quality NLEs for thoracic pathologies.
3
Song, Sihan, Chuncheng Yang, Li Xu, Haibin Shang, Zhuo Li et Yinghui Chang. « TravelRAG : A Tourist Attraction Retrieval Framework Based on Multi-Layer Knowledge Graph ». ISPRS International Journal of Geo-Information 13, no 11 (16 novembre 2024) : 414. http://dx.doi.org/10.3390/ijgi13110414.
Texte intégralRésumé :
A novel framework called TravelRAG is introduced in this paper, which is built upon a large language model (LLM) and integrates Retrieval-Augmented Generation (RAG) with knowledge graphs to create a retrieval system framework designed for the tourism domain. This framework seeks to address the challenges LLMs face in providing precise and contextually appropriate responses to domain-specific queries in the tourism field. TravelRAG extracts information related to tourist attractions from User-Generated Content (UGC) on social media platforms and organizes it into a multi-layer knowledge graph. The travel knowledge graph serves as the core retrieval source for the LLM, enhancing the accuracy of information retrieval and significantly reducing the generation of erroneous or fabricated responses, often termed as “hallucinations”. As a result, the accuracy of the LLM’s output is enhanced. Comparative analyses with traditional RAG pipelines indicate that TravelRAG significantly boosts both the retrieval efficiency and accuracy, while also greatly reducing the computational cost of model fine-tuning. The experimental results show that TravelRAG not only outperforms traditional methods in terms of retrieval accuracy but also better meets user needs for content generation.
4
Saran Raj. S et Dr. C. Meenakshi. « A Smart Legal Assistant for Indian Laws ». International Journal of Latest Technology in Engineering Management & ; Applied Science 14, no 4 (15 mai 2025) : 588–92. https://doi.org/10.51583/ijltemas.2025.140400064.
Texte intégralRésumé :
Abstract: This project proposes a chatbot system that integrates Retrieval-Augmented Generation (RAG) and a Knowledge Graph (KG) to address queries on Indian law. The system dynamically adapts to the complexity of user queries to optimize performance and accuracy. For straightforward questions, the RAG module retrieves relevant legal documents and generates concise answers. For complex, multi-faceted queries, the system employs a multihop reasoning approach using the knowledge graph to derive accurate and context- aware responses.To enhance accessibility and inclusivity, the chatbot supports multiple regional languages, enabling users from diverse linguistic backgrounds to interact with the system in their preferred language. This hybrid design ensures efficient computation by selectively engaging the KG only when necessary, reducing resource usage while maintaining high accuracy.The chatbot provides users with seamless access to legal insights, offering quick answers for simple queries and in- depth reasoning for complex cases. By combining the strengths of RAG and KG, this system aims to revolutionize legal assistance by improving accessibility, precision, and usability for both legal professionals and the general public, empowering users to navigate Indian law with confidence and efficiency.
5
Zhang, Haiyu, Yinghui Zhao, Boyu Sun, Yaqi Wu, Zetian Fu et Xinqing Xiao. « Large Language Model Based Intelligent Fault Information Retrieval System for New Energy Vehicles ». Applied Sciences 15, no 7 (6 avril 2025) : 4034. https://doi.org/10.3390/app15074034.
Texte intégralRésumé :
In recent years, the rapid development of the new energy vehicle (NEV) industry has exposed significant deficiencies in intelligent fault diagnosis and information retrieval technologies, especially in intelligent fault information retrieval, which faces persistent challenges including inadequate system adaptability and reasoning bottlenecks. To address these challenges, this study proposes a Retrieval-Augmented Generation (RAG) framework that integrates large language models (LLMs) with knowledge graphs (KGs). The framework consists of three key components: fault data collection, knowledge graph construction, and fault knowledge model training. The primary research contributions are threefold: (1) A domain-optimized fine-tuning strategy for LLMs based on NEV fault characteristics, verifying the superior accuracy of the Bidirectional Encoder Representations from Transformers (BERT) model in fault classification tasks. (2) A structured knowledge graph encompassing 122 fault categories, developed through the ChatGLM3-6B model completing named entity and knowledge relation extraction to generate fault knowledge and build a paraphrased vocabulary. (3) An intelligent fault information retrieval system that significantly outperforms traditional models in NEV-specific Q&A scenarios, providing multi-level fault cause analysis and actionable solution recommendations.
6
Martin, Andreas, Hans Friedrich Witschel, Maximilian Mandl et Mona Stockhecke. « Semantic Verification in Large Language Model-based Retrieval Augmented Generation ». Proceedings of the AAAI Symposium Series 3, no 1 (20 mai 2024) : 188–92. http://dx.doi.org/10.1609/aaaiss.v3i1.31199.
Texte intégralRésumé :
This position paper presents a novel approach of semantic verification in Large Language Model-based Retrieval Augmented Generation (LLM-RAG) systems, focusing on the critical need for factually accurate information dissemination during public debates, especially prior to plebiscites e.g. in direct democracies, particularly in the context of Switzerland. Recognizing the unique challenges posed by the current generation of Large Language Models (LLMs) in maintaining factual integrity, this research proposes an innovative solution that integrates retrieval mechanisms with enhanced semantic verification processes. The paper outlines a comprehensive methodology following a Design Science Research approach, which includes defining user personas, designing conversational interfaces, and iteratively developing a hybrid dialogue system. Central to this system is a robust semantic verification framework that leverages a knowledge graph for fact-checking and validation, ensuring the correctness and consistency of information generated by LLMs. The paper discusses the significance of this research in the context of Swiss direct democracy, where informed decision-making is pivotal. By improving the accuracy and reliability of information provided to the public, the proposed system aims to support the democratic process, enabling citizens to make well-informed decisions on complex issues. The research contributes to advancing the field of natural language processing and information retrieval, demonstrating the potential of AI and LLMs in enhancing civic engagement and democratic participation.
7
Yao, Yao, et Horacio González–Vélez. « AI-Powered System to Facilitate Personalized Adaptive Learning in Digital Transformation ». Applied Sciences 15, no 9 (30 avril 2025) : 4989. https://doi.org/10.3390/app15094989.
Texte intégralRésumé :
As Large Language Models (LLMs) incorporate generative Artificial Intelligence (AI) and complex machine learning algorithms, they have proven to be highly effective in assisting human users with complex professional tasks through natural language interaction. However, in addition to their current capabilities, LLMs occasionally generate responses that contain factual inaccuracies, stemming from their dependence on the parametric knowledge they encapsulate. To avoid such inaccuracies, also known as hallucinations, people use domain-specific knowledge (expertise) to support LLMs in the corresponding task, but the necessary knowledge engineering process usually requires considerable manual effort from experts. In this paper, we developed an approach to leverage the collective strengths of multiple agents to automatically facilitate the knowledge engineering process and then use the learned knowledge and Retrieval Augmented Generation (RAG) pipelines to optimize the performance of LLMs in domain-specific tasks. Through this approach, we effectively build AI assistants based on particular customized knowledge to help students better carry out personalized adaptive learning in digital transformation. Our initial tests demonstrated that integrating a Knowledge Graph (KG) within a RAG framework significantly improved the quality of domain-specific outputs generated by the LLMs. The results also revealed performance fluctuations for LLMs across varying contexts, underscoring the critical need for domain-specific knowledge support to enhance AI-driven adaptive learning systems.
8
Balasubramanian, Abhinav. « Accelerating Research with Automated Literature Reviews : A Rag-Based Framework ». International Journal of Multidisciplinary Research and Growth Evaluation. 6, no 2 (2025) : 337–42. https://doi.org/10.54660/.ijmrge.2025.6.2.337-342.
Texte intégralRésumé :
The exponential growth of academic publications has significantly increased the complexity of synthesizing knowledge across various disciplines. Researchers often struggle to manually analyze vast volumes of literature, a process that is both time-consuming and prone to biases. These challenges highlight the urgent need for innovative solutions that can streamline the literature review process and improve the quality of knowledge synthesis. This paper proposes a theoretical framework based on Retrieval-Augmented Generation (RAG) to automate the aggregation and summarization of academic literature. By integrating semantic search, generative AI, and knowledge graph technology, the framework offers a comprehensive solution to efficiently retrieve, synthesize, and contextualize key findings from relevant academic works. The use of knowledge graphs enhances the identification of research trends and gaps, offering researchers a deeper understanding of interconnected topics and areas requiring further exploration. Key contributions of this work include the conceptualization of the RAG-based framework and the introduction of a theoretical evaluation methodology. The evaluation metrics focus on semantic relevance, contextual coherence, source diversity, and usability, providing a robust foundation for assessing the framework’s potential. By reducing the time and effort required for literature reviews, this framework aims to accelerate innovation, facilitate interdisciplinary collaboration, and transform traditional research workflows.
9
Yang, Jiawei, Chuanyao Sun, Junwu Zhou, Qingkai Wang, Kanghui Zhang et Tao Song. « Knowledge-Inference-Based Intelligent Decision Making for Nonferrous Metal Mineral-Processing Flowsheet Design ». Minerals 15, no 4 (3 avril 2025) : 374. https://doi.org/10.3390/min15040374.
Texte intégralRésumé :
With the increasing diversification of ore types and the complexity of processing techniques in the mining industry, traditional decision-making methods for mineral processing flowsheets can no longer meet the high efficiency and intelligence requirements. This paper proposes a knowledge graph-based framework for constructing a mineral-processing design knowledge base and knowledge reasoning, aiming at providing intelligent and efficient decision support for mining engineers. This framework integrates Chinese NLP models for text vectorization, optimizes prompt generation through Retrieval Augmented Generation (RAG) technology, realizes knowledge graph construction, and implements knowledge reasoning for nonferrous metal mineral-processing design using large reasoning models. By analyzing the genetic characteristics of ores and the requirements of processing techniques, the framework outputs reasonable flowsheet designs, which could help engineers save research time and labor in optimizing processes, selecting suitable reagents, and adjusting process parameters. Through decision analysis of the mineral-processing flowsheets for three typical copper mines, the framework demonstrates its advantages in improving process flowsheet design, and shows good potential for further application in complex mineral-processing environments.
10
VanGundy, Braxton, Nipa Phojanamongkolkij, Barclay Brown, Ramana Polavarapu et Joshua Bonner. « Requirement Discovery Using Embedded Knowledge Graph with ChatGPT ». INCOSE International Symposium 34, no 1 (juillet 2024) : 2011–27. http://dx.doi.org/10.1002/iis2.13253.
Texte intégralRésumé :
AbstractThe field of Advanced Air Mobility (AAM) is witnessing a transformation with innovations such as electric aircraft and increasingly automated airspace operations. Within AAM, the Urban Air Mobility (UAM) concept focuses on providing air‐taxi services in densely populated urban areas. This research introduces the utilization of Large Language Models (LLMs), such as OpenAI's GPT‐4, to enhance the UAM Requirement discovery process.This study explores two distinct approaches to leverage LLMs in the context of UAM Requirement discovery. The first approach evaluates the LLM's ability to provide responses without relying on additional outside systems, such as a relational or graph database. Instead, a vector store provides relevant information to the LLM based on the user's question, a process known as Retrieval Augmented Generation (RAG). The second approach integrates the LLM with a graph database. The LLM acts as an intermediary between the user and the graph database, translating user questions into cypher queries for the database and database responses into human‐readable answers for the user. Our team implemented and tested both solutions to analyze requirements within a UAM dataset. This paper will talk about our approaches, implementations, and findings related to both approaches.
11
Wang, Yu, Nedim Lipka, Ryan A. Rossi, Alexa Siu, Ruiyi Zhang et Tyler Derr. « Knowledge Graph Prompting for Multi-Document Question Answering ». Proceedings of the AAAI Conference on Artificial Intelligence 38, no 17 (24 mars 2024) : 19206–14. http://dx.doi.org/10.1609/aaai.v38i17.29889.
Texte intégralRésumé :
The `pre-train, prompt, predict' paradigm of large language models (LLMs) has achieved remarkable success in open-domain question answering (OD-QA). However, few works explore this paradigm in multi-document question answering (MD-QA), a task demanding a thorough understanding of the logical associations among the contents and structures of documents. To fill this crucial gap, we propose a Knowledge Graph Prompting (KGP) method to formulate the right context in prompting LLMs for MD-QA, which consists of a graph construction module and a graph traversal module. For graph construction, we create a knowledge graph (KG) over multiple documents with nodes symbolizing passages or document structures (e.g., pages/tables), and edges denoting the semantic/lexical similarity between passages or document structural relations. For graph traversal, we design an LLM-based graph traversal agent that navigates across nodes and gathers supporting passages assisting LLMs in MD-QA. The constructed graph serves as the global ruler that regulates the transitional space among passages and reduces retrieval latency. Concurrently, the graph traversal agent acts as a local navigator that gathers pertinent context to progressively approach the question and guarantee retrieval quality. Extensive experiments underscore the efficacy of KGP for MD-QA, signifying the potential of leveraging graphs in enhancing the prompt design and retrieval augmented generation for LLMs. Our code: https://github.com/YuWVandy/KG-LLM-MDQA.
12
Sun, Yi, Ying Han et Xinke Liu. « Intelligent Gas Risk Assessment and Report Generation for Coal Mines : An Innovative Framework Based on GLM Fine-Tuning ». Electronics 14, no 2 (19 janvier 2025) : 379. https://doi.org/10.3390/electronics14020379.
Texte intégralRésumé :
Traditional coal mine gas risk assessment relies on manual operations, leading to inefficiencies, incomplete information integration, and insufficient evaluation accuracy, ultimately affecting safety oversight. This paper proposes an intelligent gas risk assessment and report generation framework (IGRARG) based on fine-tuning a Generative Language Model (GLM) to address these challenges. The framework integrates multi-source sensor data with the reasoning capabilities of large language models (LLMs). It constructs a gas risk dataset for coal mine safety scenarios, fine-tuned with GLM. Incorporating industry regulations and a domain-specific knowledge base enhanced with a Retrieval-Augmented Generation (RAG) mechanism, the framework automates alarm judgment, suggestion generation, and report creation via a hierarchical graph structure. Real-time human feedback further refines decision making. Experimental results show an evaluation accuracy of 85–93%, with over 300 field tests achieving a 94.46% alarm judgment accuracy and reducing weekly report generation from 90 min to 2–3 min. This framework significantly enhances the intelligence and efficiency of gas risk assessment, providing robust decision support for coal mine safety management.
13
Mendel, Ofek, Michael S. Sinclair, Kathleen M. Jagodnik, Till Krenz, Jeronimo Pissinis, Vasileios Stathias, Alon Bartal et Stephan Schürer. « Abstract 1052 : Integrating geographical and socioeconomic data with genomic information for enhanced prediction of cancer risk using large language models and knowledge graphs ». Cancer Research 85, no 8_Supplement_1 (21 avril 2025) : 1052. https://doi.org/10.1158/1538-7445.am2025-1052.
Texte intégralRésumé :
Cancer remains a leading cause of mortality worldwide, with risk and survival outcomes influenced by a complex interplay of genetic, environmental, and socioeconomic factors. Traditional cancer risk prediction models often focus on genetic, lifestyle, and family history factors but frequently overlook the critical roles of geographical location and social determinants of health (SDoH). This study introduces a novel approach that integrates diverse multimodal datasets—including environmental exposures, healthcare access, socioeconomic conditions, and genomic data—within a geographic framework to develop more comprehensive and spatially contextualized cancer risk prediction and survival models. First, we developed a data model to integrate the various data modalities. We leverage our internal Sylvester Data Portal Informatics Platform to manage and map the various data types and metadata. Using the GPT-4o Large Language Model, we identified entities and their relationships from semi-structured and unstructured data sources, and used them to construct a multimodal knowledge graph (KG). Our KG comprises 700,000+ nodes spanning 32 entity types (e.g., genes, 5 cancer types) and 2.134 million edges/links across 107 relationship types (e.g., has_genotype, from_race, from_sex). Graph Neural Networks (GNNs) were then trained on this KG to predict cancer risk and survival outcomes based on demographic factors, location, and tumor genetics. Querying our KG using Graph Retrieval Augmented Generation (GraphRAG) and predicting links using our trained GNN, we uncovered novel insights into the relationships among genetics, demographics, environmental exposures, and cancer statistics, highlighting potential complex gene-environment interactions that contribute to cancer risk and survival disparities, and suggesting directions for future research to reduce cancer risk and improve health outcomes. Citation Format: Ofek Mendel, Michael S. Sinclair, Kathleen M. Jagodnik, Till Krenz, Jeronimo Pissinis, Vasileios Stathias, Alon Bartal, Stephan Schürer. Integrating geographical and socioeconomic data with genomic information for enhanced prediction of cancer risk using large language models and knowledge graphs [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 1052.
14
Long, Xinwei, Zhiyuan Ma, Ermo Hua, Kaiyan Zhang, Biqing Qi et Bowen Zhou. « Retrieval-Augmented Visual Question Answering via Built-in Autoregressive Search Engines ». Proceedings of the AAAI Conference on Artificial Intelligence 39, no 23 (11 avril 2025) : 24723–31. https://doi.org/10.1609/aaai.v39i23.34653.
Texte intégralRésumé :
Retrieval-augmented generation (RAG) has emerged to address the knowledge-intensive visual question answering (VQA) task. Current methods mainly employ separate retrieval and generation modules to acquire external knowledge and generate answers, respectively. We propose ReAuSE, an alternative to the previous RAG model for the knowledge-based VQA task, which seamlessly integrates knowledge retriever into the generative multi-modal large language model, serving as a built-in search engine. Specifically, our model functions both as a generative retriever and an accurate answer generator. It not only helps retrieve documents from the knowledge base by producing identifier for each document, but it also answers visual questions based on the retrieved documents. Furthermore, we also propose a reinforced retrieval calibration module from relevance feedback to improve retrieval performance and align with the preferences for accurate answer generation. Extensive experiments on two representative OKVQA and A-OKVQA datasets demonstrate significant improvements ranging from 2.9% to 9.6% across all evaluation metrics when compared to strong baselines.
15
Si, Han, Sanyam Kumar, Sneh Lata, Arshad Ahmad, Saurav Ghosh, Karen Stephansen, Deepti Nagarkar, Eda Zhou et Brandon W. Higgs. « Abstract 3644 : Mechanistically explainable AI model for predicting synergistic cancer therapy combinations ». Cancer Research 85, no 8_Supplement_1 (21 avril 2025) : 3644. https://doi.org/10.1158/1538-7445.am2025-3644.
Texte intégralRésumé :
Abstract Background: Resistance to single-agent cancer therapies often arises through complex mechanisms such as genetic mutations, compensatory signaling, and tumor microenvironment changes. These adaptive strategies make multi-drug combinations essential to effectively target canonical pathways simultaneously. However, common in vitro and in vivo models often fail to capture these complexities, resulting in limited translational success to human clinical settings. This highlights the need for predictive frameworks that not only identify synergistic drug combinations from human clinical trials, but also provide mechanistic insights to inform clinical application. This study presents a novel Large Language Model (LLM)-based framework that integrates extensive drug combination data with a knowledge graph to predict drug combinations with biological explainability. Methods: A retrieval-augmented generation (RAG) approach was employed to enhance predictive specificity and reduce hallucinations. In vitro data on drugs, targets, cell lines, tumor types, and synergy scores were compiled from over 50, 000 cell line results from the DrugComboDb database. Using NCBI’s APIs, BeautifulSoup, and GPT-4, biomedical information was extracted from 1, 092 phase I-III clinical trials covering 700+ drug combinations across 22 tumor types. This RAG corpus, combined with targeted prompts, fed into the LLM Mistral v0.2 for prediction generation, while a knowledge graph - PrimeKG provided mechanistic insights. Model validation included unseen cases (N=42) across various drug modality combinations (N=8), including 34 immuno-oncology therapies (IO) combined with an IO, chemo, or targeted therapy, 1 targeted therapy combination and 7 antibody drug conjugate (ADC) combinations. Results: The model achieved an F1 score of 0.80 on the entire validation set, with accuracy at 0.74, precision at 0.71, and recall at 0.92. Among drug modalities, combinations of ADC agents with IO/targeted/chemo/other ADC agents performed best (F1 score: 100%), while IO-IO combinations scored the lowest (F1 score: 62%). Each prediction included detailed mechanistic explanations, highlighting whether the drug combination was supported or unsupported with biological rationale. Conclusions: This study presents a novel framework integrating LLMs with biological knowledge graphs to predict the efficacy when combining 2-3 oncology drugs across a range of indications. By linking generative AI with explainability tools, we address two critical challenges in AI-driven drug discovery: predictive accuracy and mechanistic transparency. Such predictive models can accelerate the discovery of synergistic therapies, thereby reducing costly experiments, and improve overall patient outcomes. Specific test cases with explainable results will be presented. Citation Format: Han Si, Sanyam Kumar, Sneh Lata, Arshad Ahmad, Saurav Ghosh, Karen Stephansen, Deepti Nagarkar, Eda Zhou, Brandon W. Higgs. Mechanistically explainable AI model for predicting synergistic cancer therapy combinations [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3644.
16
Liu, Yicheng. « Retrieval-Augmented Generation : Methods, Applications and Challenges ». Applied and Computational Engineering 142, no 1 (24 avril 2025) : 99–108. https://doi.org/10.54254/2755-2721/2025.kl22312.
Texte intégralRésumé :
The Retrieval-Augmented Generation (RAG) has been proven to have a promising approach. It can address the limitations of purely generative models in knowledge-intensive tasks caused by their reliance on static, pre-trained knowledge. RAG addresses these challenges by integrating a retrieval mechanism with a generative model, enabling dynamic access to external knowledge sources during the generation process. This paper presents a comprehensive study of the RAG framework, focusing on its architecture, training strategies, and applications. The framework combines a dense passage retriever (DPR) with a sequence-to-sequence generator (GPT-3.5-turbo), jointly optimized in an end-to-end manner to retrieve and utilize relevant knowledge effectively. This paper evaluates RAG on MS MARCO, demonstrating its superiority over state-of-the-art purely generative models and traditional retrieval-based systems. Experimental results show that RAG achieves significant improvements in factual accuracy, relevance, and interpretability, as measured by metrics such as term frequencyinverse document frequency, bidirectional encoder representation from transformer Score, and Q-Bilingual Evaluation Understudy-1.
17
Zhu, Xishi, Xiaoming Guo, Shengting Cao, Shenglin Li et Jiaqi Gong. « StructuGraphRAG : Structured Document-Informed Knowledge Graphs for Retrieval-Augmented Generation ». Proceedings of the AAAI Symposium Series 4, no 1 (8 novembre 2024) : 242–51. http://dx.doi.org/10.1609/aaaiss.v4i1.31798.
Texte intégralRésumé :
Retrieval-augmented generation (RAG) enhances large language models (LLMs) by incorporating external data sources beyond their training sets and querying predefined knowledge bases to generate accurate, context-rich responses. Most RAG implementations use vector similarity searches, but the effectiveness of this approach and the representation of knowledge bases remain underexplored. Emerging research suggests knowledge graphs as a promising solution. Therefore, this paper presents StructuGraphRAG, which leverages document structures to inform the extraction process and constructs knowledge graphs to enhance RAG for social science research, specifically using NSDUH datasets. Our method parses document structures to extract entities and relationships, constructing comprehensive and relevant knowledge graphs. Experimental results show that StructuGraphRAG outperforms traditional RAG methods in accuracy, comprehensiveness, and contextual relevance. This approach provides a robust tool for social science researchers, facilitating precise analysis of social determinants of health and justice, and underscores the potential of structured document-informed knowledge graph construction in AI and social science research.
18
Zhu, Jia, Hanghui Guo, Weijie Shi, Zhangze Chen et Pasquale De Meo. « RaDIO : Real-Time Hallucination Detection with Contextual Index Optimized Query Formulation for Dynamic Retrieval Augmented Generation ». Proceedings of the AAAI Conference on Artificial Intelligence 39, no 24 (11 avril 2025) : 26129–37. https://doi.org/10.1609/aaai.v39i24.34809.
Texte intégralRésumé :
The Dynamic Retrieval Augmented Generation (RAG) paradigm actively decides when and what to retrieve during the text generation process of Large Language Models (LLMs). However, current dynamic RAG methods fall short in both aspects: identifying the optimal moment to activate the retrieval module and crafting the appropriate query once retrieval is triggered. To overcome these limitations, we introduce an approach, namely, RaDIO, Real-Time Hallucination Detection with Contextual Index Optimized query formulation for dynamic RAG. The approach is specifically designed to make decisions on when and what to retrieve based on the LLM’s real-time information needs during the text generation process. We evaluate RaDIO along with existing methods comprehensively over several knowledge-intensive generation datasets. Experimental results show that RaDIO achieves superior performance on all tasks, demonstrating the effectiveness of our work.
19
Selinger, Douglas W., Timothy R. Wall, Eleni Stylianou, Ehab Khalil, Jedidiah Gaetz et Oren Levy. « Abstract B043 : A fully transparent and automatable form of AI for multi-omics precision medicine ». Cancer Research 85, no 5_Supplement (11 mars 2025) : B043. https://doi.org/10.1158/1538-7445.genfunc25-b043.
Texte intégralRésumé :
Abstract Acquired drug resistance and heterogeneity of patient responses to cancer therapeutics represent major challenges to delivering effective treatments to patients. Predictive biomarkers provide valuable information for increasing drug responses among treatment cohorts by guiding the delivery of drugs to patients most likely to benefit. However, the identification of biomarkers of drug sensitivity and resistance remains challenging, and their mechanistic basis is difficult to discern. Large-scale multi-omics methods represent a promising approach; however, they capture far more data than can be digested in human readable form, creating an opportunity for machine-based technologies to significantly expand the boundaries of scientific knowledge. Artificial intelligence (AI) methods promise to fill this gap, however significant challenges remain to adapt machine learning (ML) based approaches to data that is noisy, complex, and bespoke while still preserving the transparency required for scientific and regulatory decision making. Herein we describe a novel approach, which combines knowledge graphs and centrality algorithms, allowing the construction of systems that are scalable, robust to noise, concise, and perhaps most importantly, highly transparent. The central construct, which we’ve termed a “focal graph”, can seamlessly integrate information across multiple, diverse, complex data sets, including large-scale chemical biology and multi-omics data. Focal graphs can be combined with large language models (LLMs) as part of a retrieval-augmented generation (RAG) strategy to build intelligent, autonomous drug discovery workflows that preserve the provenance of the underlying experimental data, and whose processes and conclusions can be examined and evaluated in tremendous detail. Focal graphs can be paired with LLMs to autonomously plan and execute research programs, potentially resulting in discoveries that are entirely novel, and derive their support from multiple, diverse, large-scale experimental data sets. As such autonomous systems become more sophisticated and are given access to more data and computational power, they can be expected to generate insights with increasing levels of novelty and experimental support. Here we present the theoretical underpinnings of focal graphs and their automation, as well as selected examples highlighting their application to biomarker development and the investigation of mechanisms underlying the heterogeneity of therapeutic responses in cancer. More information is available at www.plexresearch.com. Citation Format: Douglas W Selinger, Timothy R Wall, Eleni Stylianou, Ehab Khalil, Jedidiah Gaetz, Oren Levy. A fully transparent and automatable form of AI for multi-omics precision medicine [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Functional and Genomic Precision Medicine in Cancer: Different Perspectives, Common Goals; 2025 Mar 11-13; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(5 Suppl):Abstract nr B043.
20
Grabuloski, Marko, Aleksandar Karadimce et Anis Sefidanoski. « Enhancing Language Models with Retrieval-Augmented Generation A Comparative Study on Performance ». WSEAS TRANSACTIONS ON INFORMATION SCIENCE AND APPLICATIONS 22 (2 avril 2025) : 272–97. https://doi.org/10.37394/23209.2025.22.23.
Texte intégralRésumé :
Retrieval-Augmented Generation (RAG) is a powerful technique that enhances the capabilities of Large Language Models (LLMs) by integrating information retrieval with text generation. By accessing and incorporating relevant external knowledge, RAG systems address the limitations of traditional LLMs, such as memory constraints and the inability to access up-to-date information. This research explores the implementation and evaluation of RAG systems, focusing on their potential to improve the accuracy and relevance of LLM responses. It investigates the impact of different LLM types (causal, question-answering, conversational) and retrieval-augmentation strategies (sentence-level, paragraph-level) on the performance of RAG systems. We conducted experiments using various open-source LLMs and a custom-built RAG system to assess the effectiveness of different approaches. The findings indicate that RAG systems can significantly enhance the performance of LLMs, especially for complex questions that require access to diverse information sources. T5 conversational models, in particular, demonstrate strong performance in synthesis-based tasks, effectively combining information from multiple retrieved documents. However, causal and question-answering models may struggle with complex reasoning and synthesis, even with RAG augmentation.
21
Ajay Mukund, S., et K. S. Easwarakumar. « Optimizing Legal Text Summarization Through Dynamic Retrieval-Augmented Generation and Domain-Specific Adaptation ». Symmetry 17, no 5 (23 avril 2025) : 633. https://doi.org/10.3390/sym17050633.
Texte intégralRésumé :
Legal text summarization presents distinct challenges due to the intricate and domain-specific nature of legal language. This paper introduces a novel framework integrating dynamic Retrieval-Augmented Generation (RAG) with domain-specific adaptation to enhance the accuracy and contextual relevance of legal document summaries. The proposed Dynamic Legal RAG system achieves a vital form of symmetry between information retrieval and content generation, ensuring that retrieved legal knowledge is both comprehensive and precise. Using the BM25 retriever with top-3 chunk selection, the system optimizes relevance and efficiency, minimizing redundancy while maximizing legally pertinent content. with top-3 chunk selection, the system optimizes relevance and efficiency, minimizing redundancy while maximizing legally pertinent content. A key design feature is the compression ratio constraint (0.05 to 0.5), maintaining structural symmetry between the original judgment and its summary by balancing representation and information density. Extensive evaluations establish BM25 as the most effective retriever, striking an optimal balance between precision and recall. A comparative analysis of transformer-based (Decoder-only) models—DeepSeek-7B, LLaMA 2-7B, and LLaMA 3.1-8B—demonstrates that LLaMA 3.1-8B, enriched with Legal Named Entity Recognition (NER) and the Dynamic RAG system, achieves superior performance with a BERTScore of 0.89. This study lays a strong foundation for future research in hybrid retrieval models, adaptive chunking strategies, and legal-specific evaluation metrics, with practical implications for case law analysis and automated legal drafting.
22
Han, Binglan, Teo Susnjak et Anuradha Mathrani. « Automating Systematic Literature Reviews with Retrieval-Augmented Generation : A Comprehensive Overview ». Applied Sciences 14, no 19 (9 octobre 2024) : 9103. http://dx.doi.org/10.3390/app14199103.
Texte intégralRésumé :
This study examines Retrieval-Augmented Generation (RAG) in large language models (LLMs) and their significant application for undertaking systematic literature reviews (SLRs). RAG-based LLMs can potentially automate tasks like data extraction, summarization, and trend identification. However, while LLMs are exceptionally proficient in generating human-like text and interpreting complex linguistic nuances, their dependence on static, pre-trained knowledge can result in inaccuracies and hallucinations. RAG mitigates these limitations by integrating LLMs’ generative capabilities with the precision of real-time information retrieval. We review in detail the three key processes of the RAG framework—retrieval, augmentation, and generation. We then discuss applications of RAG-based LLMs to SLR automation and highlight future research topics, including integration of domain-specific LLMs, multimodal data processing and generation, and utilization of multiple retrieval sources. We propose a framework of RAG-based LLMs for automating SRLs, which covers four stages of SLR process: literature search, literature screening, data extraction, and information synthesis. Future research aims to optimize the interaction between LLM selection, training strategies, RAG techniques, and prompt engineering to implement the proposed framework, with particular emphasis on the retrieval of information from individual scientific papers and the integration of these data to produce outputs addressing various aspects such as current status, existing gaps, and emerging trends.
23
Wang, Haitao, et Yangkun Shi. « Knowledge Graph Combined with Retrieval-Augmented Generation for Enhancing LMs Reasoning : A Survey ». Academic Journal of Science and Technology 14, no 1 (12 février 2025) : 227–35. https://doi.org/10.54097/h21fky45.
Texte intégralRésumé :
Large language models (LLMs) have generated significant waves in the fields of Natural Language Processing(NLP) and artificial intelligence due to their remarkable capabilities and broad adaptability. Retrieval-Augmented Generation (RAG) techniques have been widely adopted, leveraging external retrieval systems to significantly improve the timeliness of LLMs and substantially reduce hallucinations. Although RAG and its optimization methods have addressed most hallucination issues caused by knowledge gaps and outdated information, text generation in specialized domains such as law, medicine, and science—which require multi-hop reasoning and analysis—still suffers from a lack of coherence and logical consistency, making it difficult to produce correct and valuable answers. To address these challenges, related research has introduced Knowledge Graphs (KGs). This integrated approach enhances RAG’s capabilities by utilizing the structured knowledge provided by KGs, thereby improving the model's knowledge representation and reasoning abilities and enabling the generation of more accurate answers. However, there remains a lack of systematic reviews in this area. Therefore, this paper provides a comprehensive review of studies on enhancing LLMs reasoning abilities by integrating KGs with RAG. It first introduces the basic concepts, followed by an overview of the current mainstream technical approaches, and concludes with a discussion of the research challenges and future development trends in this field.
24
Xu, Kehan, Kun Zhang, Jingyuan Li, Wei Huang et Yuanzhuo Wang. « CRP-RAG : A Retrieval-Augmented Generation Framework for Supporting Complex Logical Reasoning and Knowledge Planning ». Electronics 14, no 1 (26 décembre 2024) : 47. https://doi.org/10.3390/electronics14010047.
Texte intégralRésumé :
The Retrieval-Augmented Generation (RAG) framework enhances Large Language Models (LLMs) by retrieving relevant knowledge to broaden their knowledge boundaries and mitigate factual hallucinations stemming from knowledge gaps. However, the RAG Framework faces challenges in effective knowledge retrieval and utilization; invalid or misused knowledge will interfere with LLM generation, reducing reasoning efficiency and answer quality. Existing RAG methods address these issues by decomposing and expanding queries, introducing special knowledge structures, and using reasoning process evaluation and feedback. However, the linear reasoning structures limit complex thought transformations and reasoning based on intricate queries. Additionally, knowledge retrieval and utilization are decoupled from reasoning and answer generation, hindering effective knowledge support during answer generation. To address these limitations, we propose the CRP-RAG framework, which employs reasoning graphs to model complex query reasoning processes more comprehensively and accurately. CRP-RAG guides knowledge retrieval, aggregation, and evaluation through reasoning graphs, dynamically adjusting the reasoning path based on evaluation results and selecting knowledge-sufficiency paths for answer generation. CRP-RAG outperforms the best LLM and RAG baselines by 2.46 in open-domain QA, 7.43 in multi-hop reasoning, and 4.2 in factual verification. Experiments also show the superior factual consistency and robustness of CRP-RAG over existing RAG methods. Extensive analyses confirm its accurate and fact-faithful reasoning and answer generation for complex queries.
25
Siriwardhana, Shamane, Rivindu Weerasekera, Elliott Wen, Tharindu Kaluarachchi, Rajib Rana et Suranga Nanayakkara. « Improving the Domain Adaptation of Retrieval Augmented Generation (RAG) Models for Open Domain Question Answering ». Transactions of the Association for Computational Linguistics 11 (2023) : 1–17. http://dx.doi.org/10.1162/tacl_a_00530.
Texte intégralRésumé :
Abstract Retrieval Augment Generation (RAG) is a recent advancement in Open-Domain Question Answering (ODQA). RAG has only been trained and explored with a Wikipedia-based external knowledge base and is not optimized for use in other specialized domains such as healthcare and news. In this paper, we evaluate the impact of joint training of the retriever and generator components of RAG for the task of domain adaptation in ODQA. We propose RAG-end2end, an extension to RAG that can adapt to a domain-specific knowledge base by updating all components of the external knowledge base during training. In addition, we introduce an auxiliary training signal to inject more domain-specific knowledge. This auxiliary signal forces RAG-end2end to reconstruct a given sentence by accessing the relevant information from the external knowledge base. Our novel contribution is that, unlike RAG, RAG-end2end does joint training of the retriever and generator for the end QA task and domain adaptation. We evaluate our approach with datasets from three domains: COVID-19, News, and Conversations, and achieve significant performance improvements compared to the original RAG model. Our work has been open-sourced through the HuggingFace Transformers library, attesting to our work’s credibility and technical consistency.
26
Yang, Chi Bok, et Yang Sok Kim. « Implementation of Retrieval Augmented Generation (RAG) Model Using LLM : A RapidMiner-Based Approach ». Korean Institute of Smart Media 14, no 2 (28 février 2025) : 34–42. https://doi.org/10.30693/smj.2025.14.2.34.
Texte intégralRésumé :
Generative AI technology, driven by Large Language Models (LLMs), is being increasingly utilized to overcome existing limitations. Retrieval-Augmented Generation (RAG) has emerged as an effective approach to reduce hallucination in LLMs by leveraging up-to-date and domain-specific knowledge beyond training data. However, most studies propose programming-based implementations. This research introduces a GUI-based RAG framework using RapidMiner, to construct RAG systems without programming proficiency. The methodology includes storing and retrieving embeddings with the Qdrant vector database and generating question-and-answer pairs via the OpenAI API. Practical demonstrations confirm the system’s effectiveness in real-world scenarios, offering a simpler and more efficient method for developing generative AI services with LLMs.
27
Lin, Xin, Zhenya Huang, Zhiqiang Zhang, Jun Zhou et Enhong Chen. « Explore What LLM Does Not Know in Complex Question Answering ». Proceedings of the AAAI Conference on Artificial Intelligence 39, no 23 (11 avril 2025) : 24585–94. https://doi.org/10.1609/aaai.v39i23.34638.
Texte intégralRésumé :
Complex question answering (QA) is a challenging task in artificial intelligence research which requires reasoning based on related knowledge. The retrieval-augmented generation (RAG) based on large language models (LLMs) have become one promising solution in QA. To facilitate RAG more effectively, the LLM needs to precisely evaluate knowledge required in QA. That is, first, the LLM needs to examine its knowledge boundary (what the LLM does not know) to retrieve external knowledge as supplement. Second, the LLM needs to evaluate the utility of the retrieved knowledge (whether it helps in reasoning) for robust RAG. To this end, in this paper, we propose a novel Question Answering with Knowledge Evaluation (KEQA) framework to promote the effectiveness and efficiency of RAG in QA. First, inspired by quizzes in classroom, we propose a quiz-based method to precisely examine the knowledge state of the uninterpretable LLM for QA. We ask indicative quizzes on each required knowledge, and inspect whether the LLM can consistently answer the quiz to examine its knowledge boundary. Second, we retrieve the unknown knowledge from external source, and evaluate its utility to pick the helpful ones for reasoning. We design a reasoning-based metric to evaluate utility, and construct a demonstration set in training data for reference to guide knowledge picking in inference. We conduct extensive experiments on four widely-used QA datasets, and the results demonstrate the effectiveness of the proposed method.
28
Liu, Shuxin. « Exploring Optimal Prefetching Sizes in RaLMSpec to Enhance Retrieval-Augmented Generation Efficiency ». Highlights in Science, Engineering and Technology 138 (11 mai 2025) : 24–31. https://doi.org/10.54097/hhff9g78.
Texte intégralRésumé :
Retrieval-augmented generation (RAG) frameworks like RaLMSpec enhance language model performance by integrating external knowledge. A key method to accelerate RaLMSpec efficiency is the prefetching, which determines the number of documents to retrieve in advance to balance retrieval speed and cache utilization. This study introduces and tests both static and dynamic prefetching strategies to optimize performance in RaLMSpec. Static prefetching uses fixed sizes, while dynamic prefetching adjusts based on real-time factors including task complexity, cache hit rates, and retrieval latency. Experiments across multiple datasets, retrievers, and language models demonstrate that dynamic prefetching significantly reduces latency by 18% on average, outperforming static strategies. Dynamic prefetching adapts to varying task demands, providing better balance between retrieval and caching efficiency. Among static strategies, a prefetch size of 64 offers the best trade-off between latency reduction and cache utilization. The results highlight that dynamic prefetching is optimal for environments with fluctuating task complexity, while static prefetching with a size of 64 is effective for predictable tasks. This study provides valuable insights for improving RAG system efficiency and suggests future directions, including machine learning-based adaptations and hardware optimizations.
29
Siddharth Nandagopal. « Securing Retrieval-Augmented Generation Pipelines : A Comprehensive Framework ». Journal of Computer Science and Technology Studies 7, no 1 (12 janvier 2025) : 17–29. https://doi.org/10.32996/jcsts.2025.7.1.2.
Texte intégralRésumé :
Retrieval-Augmented Generation (RAG) has significantly enhanced the capabilities of Large Language Models (LLMs) by enabling them to access and incorporate external knowledge sources, thereby improving response accuracy and relevance. However, the security of RAG pipelines remains a paramount concern as these systems become integral to various critical applications. This paper introduces a comprehensive framework designed to secure RAG pipelines through the integration of advanced encryption techniques, zero-trust architecture, and structured guardrails. The framework employs symmetric and asymmetric encryption to protect data at rest and in transit, ensuring confidentiality and integrity throughout the data lifecycle. Adopting zero-trust principles, the framework mandates continuous verification of all entities within the data flow, effectively mitigating unauthorized access and lateral movement risks. Additionally, the implementation of guardrails, such as immutable system prompts and salted sequence tagging, fortifies the system against prompt injection and other malicious attacks. A detailed lifecycle security continuum is presented, illustrating the application of these security measures from data ingestion to decommissioning. Case studies across healthcare, finance, retail, and education sectors demonstrate the framework’s effectiveness in maintaining high performance and scalability without compromising security. This work provides a foundational model for future research and practical implementation, emphasizing the necessity of robust security protocols in the deployment of RAG-based applications.
30
Swacha, Jakub, et Michał Gracel. « Retrieval-Augmented Generation (RAG) Chatbots for Education : A Survey of Applications ». Applied Sciences 15, no 8 (11 avril 2025) : 4234. https://doi.org/10.3390/app15084234.
Texte intégralRésumé :
Retrieval-Augmented Generation (RAG) overcomes the main barrier for the adoption of LLM-based chatbots in education: hallucinations. The uncomplicated architecture of RAG chatbots makes it relatively easy to implement chatbots that serve specific purposes and thus are capable of addressing various needs in the educational domain. With five years having passed since the introduction of RAG, the time has come to check the progress attained in its adoption in education. This paper identifies 47 papers dedicated to RAG chatbots’ uses for various kinds of educational purposes, which are analyzed in terms of their character, the target of the support provided by the chatbots, the thematic scope of the knowledge accessible via the chatbots, the underlying large language model, and the character of their evaluation.
31
Loumachi, Fatma Yasmine, Mohamed Chahine Ghanem et Mohamed Amine Ferrag. « Advancing Cyber Incident Timeline Analysis Through Retrieval-Augmented Generation and Large Language Models ». Computers 14, no 2 (13 février 2025) : 67. https://doi.org/10.3390/computers14020067.
Texte intégralRésumé :
Cyber timeline analysis or forensic timeline analysis is critical in digital forensics and incident response (DFIR) investigations. It involves examining artefacts and events—particularly their timestamps and associated metadata—to detect anomalies, establish correlations, and reconstruct a detailed sequence of the incident. Traditional approaches rely on processing structured artefacts, such as logs and filesystem metadata, using multiple specialised tools for evidence identification, feature extraction, and timeline reconstruction. This paper introduces an innovative framework, GenDFIR, a context-specific approach powered via large language model (LLM) capabilities. Specifically, it proposes the use of Llama 3.1 8B in zero-shot, selected for its ability to understand cyber threat nuances, integrated with a retrieval-augmented generation (RAG) agent. Our approach comprises two main stages: (1) Data preprocessing and structuring: incident events, represented as textual data, are transformed into a well-structured document, forming a comprehensive knowledge base of the incident. (2) Context retrieval and semantic enrichment: a RAG agent retrieves relevant incident events from the knowledge base based on user prompts. The LLM processes the pertinent retrieved context, enabling a detailed interpretation and semantic enhancement. The proposed framework was tested on synthetic cyber incident events in a controlled environment, with results assessed using DFIR-tailored, context-specific metrics designed to evaluate the framework’s performance, reliability, and robustness, supported by human evaluation to validate the accuracy and reliability of the outcomes. Our findings demonstrate the practical power of LLMs in advancing the automation of cyber-incident timeline analysis, a subfield within DFIR. This research also highlights the potential of generative AI, particularly LLMs, and opens new possibilities for advanced threat detection and incident reconstruction.
32
Huang, Jie, Mo Wang, Yunpeng Cui, Juan Liu, Li Chen, Ting Wang, Huan Li et Jinming Wu. « Layered Query Retrieval : An Adaptive Framework for Retrieval-Augmented Generation in Complex Question Answering for Large Language Models ». Applied Sciences 14, no 23 (27 novembre 2024) : 11014. http://dx.doi.org/10.3390/app142311014.
Texte intégralRésumé :
Retrieval-augmented generation (RAG) addresses the problem of knowledge cutoff and overcomes the inherent limitations of pre-trained language models by retrieving relevant information in real time. However, challenges related to efficiency and accuracy persist in current RAG strategies. A key issue is how to select appropriate methods for user queries of varying complexity dynamically. This study introduces a novel adaptive retrieval-augmented generation framework termed Layered Query Retrieval (LQR). The LQR framework focuses on query complexity classification, retrieval strategies, and relevance analysis, utilizing a custom-built training dataset to train smaller models that aid the large language model (LLM) in efficiently retrieving relevant information. A central technique in LQR is a semantic rule-based approach to distinguish between different levels of multi-hop queries. The process begins by parsing the user’s query for keywords, followed by a keyword-based document retrieval. Subsequently, we employ a natural language inference (NLI) model to assess whether the retrieved document is relevant to the query. We validated our approach on multiple single-hop and multi-hop datasets, demonstrating significant improvements in both accuracy and efficiency compared to existing single-step, multi-step, and adaptive methods. Our method exhibits high accuracy and efficiency, particularly on the HotpotQA dataset, where it outperforms the Adaptive-RAG method by improving accuracy by 9.4% and the F1 score by 16.14%. The proposed approach carefully balances retrieval efficiency with the accuracy of the LLM’s responses.
33
Lee, Kyung-Yul, et Juho Bai. « PAI-NET : Retrieval-Augmented Generation Patent Network Using Prior Art Information ». Systems 13, no 4 (7 avril 2025) : 259. https://doi.org/10.3390/systems13040259.
Texte intégralRésumé :
Similar patent document retrieval is an essential task that reduces the scope of patent claimants’ searches, and numerous studies have attempted to provide automated patent search services. Recently, Retrieval-Augmented Generation (RAG) based on generative language models has emerged as an excellent method for accessing and utilizing patent knowledge environments. RAG-based patent search services offer enhanced retrieval ranking performance as AI search services by providing document knowledge similar to queries. However, achieving optimal similarity-based document ranking in search services remains a challenging task, as search methods based on document similarity do not adequately address the characteristics of patent documents. Unlike general document retrieval, the similarity of patent documents must take into account prior art relationships. To address this issue, we propose PAI-NET, a deep neural network for computing patent document similarities by incorporating expert knowledge of prior art relationships. We demonstrate that our proposed method outperforms current state-of-the-art models in patent document classification tasks through semantic distance evaluation on the USPD and KPRIS datasets. PAI-NET presents similar document candidates, demonstrating a superior patent search performance improvement of 15% over state-of-the-art methods.
34
Hong, Jiyun, Ah-Rim Joo et Sooyoung Cheon. « Strategies for Enhancing Explainability in Financial Underwriting Using RAG and LLM ». Korean Data Analysis Society 27, no 2 (30 avril 2025) : 465–76. https://doi.org/10.37727/jkdas.2025.27.2.465.
Texte intégralRésumé :
Recent advances in large language models (LLMs) have significantly expanded the capabilities of natural language processing systems. In the financial domain, delivering explainable and reliable personalised feedback has become increasingly important. However, traditional rule-based loan screening systems fall short in reflecting an individual’s financial context and providing actionable guidance. To address these limitations, this study proposes an LLM based loan screening system enhanced with RAG (retrieval augmented generation). The system retrieves relevant financial documents via vector-based search and feeds the context into a generative model. By leveraging external knowledge in real time, the system offers document-grounded, personalised financial recommendations. Experiments using real customer data show that the proposed RAG-based system significantly outperforms existing methods in both information retrieval and language generation. These results demonstrate the system’s potential to enhance the reliability and effectiveness of personalised financial services through the integration of professional document references.
35
Sarat Kiran. « Hybrid Retrieval-Augmented Generation (RAG) Systems with Embedding Vector Databases ». International Journal of Scientific Research in Computer Science, Engineering and Information Technology 11, no 2 (28 mars 2025) : 2694–702. https://doi.org/10.32628/cseit25112702.
Texte intégralRésumé :
This article explores the integration of embedding vector databases into Retrieval-Augmented Generation (RAG) systems to enhance the capabilities of large language models. The article explores how hybrid retrieval strategies combining dense vector search with traditional keyword-based methods can address the limitations of standalone LLMs, particularly regarding knowledge cutoff, hallucinations, and access to domain-specific information. The article presents a comprehensive framework covering theoretical foundations, methodological approaches, implementation considerations, and experimental results across multiple domains. By leveraging vector embeddings for semantic search alongside traditional retrieval techniques, the proposed system demonstrates significant improvements in accuracy, relevance, and factual correctness while maintaining reasonable query response times. The article provides valuable insights for enterprise-scale deployments of RAG systems across various application domains including healthcare, legal, technical support, and financial services.
36
Mahajan, Vaibhav Fanindra. « The Evolution of AI Support : How RAG is Transforming Customer Experience ». European Journal of Computer Science and Information Technology 13, no 14 (15 avril 2025) : 115–26. https://doi.org/10.37745/ejcsit.2013/vol13n14115126.
Texte intégralRésumé :
This article examines how Retrieval-Augmented Generation (RAG) is transforming customer support operations by addressing the fundamental limitations of traditional AI chatbots. While conventional chatbots rely on either rule-based systems or limited machine learning models with static knowledge bases, RAG represents a paradigm shift by dynamically retrieving information from enterprise knowledge sources before generating responses. This hybrid approach combines the strengths of retrieval-based and generation-based methods to deliver more accurate, contextually appropriate, and up-to-date support experiences. The article explores RAG's key advantages, including enhanced accuracy with reduced hallucinations, dynamic knowledge integration without manual updates, improved contextual understanding across multi-turn conversations, superior handling of complex queries, and seamless knowledge transfer to human agents when necessary. Implementation considerations covering data quality requirements, integration complexity, computational resource demands, and privacy concerns are discussed alongside real-world impact assessments and emerging future directions such as multimodal capabilities, personalized knowledge bases, proactive support models, and cross-lingual functionality. The transformative potential of RAG for customer experience represents a significant advancement in how businesses can leverage artificial intelligence to enhance support operations while reducing maintenance burdens.
37
Ms. Reshma Owhal, Viraj Shewale, Aniket Sorate, Mayur Swami et Dipak Waghmode. « CYBERVIDYA : Rag Infused Cyber Solutions ». International Research Journal on Advanced Engineering Hub (IRJAEH) 3, no 03 (15 mars 2025) : 456–64. https://doi.org/10.47392/irjaeh.2025.0063.
Texte intégralRésumé :
As cyber threats grow more sophisticated, the need for intelligent, adaptive security solutions has never been greater. CyberVidya: RAG-Infused Cyber Solutions offers a groundbreaking approach by integrating large language models (LLMs) with retrieval-augmented generation (RAG) to provide precise, real-time cybersecurity insights. Unlike traditional models that rely on static knowledge, CyberVidya continuously retrieves and processes information from a dynamic, indexed database of academic papers, ethical books, PDFs, and real-world case studies. What sets CyberVidya apart is its Non-Parametric Knowledge Retrieval, which ensures that responses are contextually accurate and directly sourced from trusted materials. Its multidimensional query-optimized retrievers work alongside advanced LLMs—GPT-2, Mistral-7B, and Llama 3.2-3B—to generate reliable, actionable insights. By incorporating document embedding and Dense Passage Retrieval (DPR), CyberVidya enhances accuracy while adapting to the ever-changing cybersecurity landscape without the need for retraining. The results speak for themselves. CyberVidya consistently outperforms industry-leading models, achieving 92.86% relevance, 85.81% similarity, and 95.06% correctness in educational queries. For scenario-based cybersecurity challenges, it maintains high performance with 92.89% relevance, 89.56% similarity, and 93.94% correctness. Comparative studies further highlight that RAG-based models surpass traditional LLMs in understanding complex cybersecurity concepts such as tactics, techniques, and procedures (TTPs) With its ability to provide accurate, real-time cybersecurity guidance, CyberVidya stands as a powerful tool for individuals, enterprises, and educators, bridging the gap between static knowledge and dynamic problem-solving.
38
Sundara Rao, Dr T. Syam, P. Venkata Snehalatha, R. Lakshmi Naga Chaitanya, M. Aparna et Sk Haleema Kusum. « RAG -CLONE A Generic Framework ». International Scientific Journal of Engineering and Management 04, no 03 (5 mars 2025) : 1–6. https://doi.org/10.55041/isjem02315.
Texte intégralRésumé :
This paper presents a RAG system (Retrieval Augmented Generation)that aims to improve how AI processes, accesses ,and generates information. Our approach uses Vector embedding to improve data absorption and provide more accurate and contextual answers using FAISS-based on the similarity and the mistral searches. The system is created to process unstructured, unstructured data from a variety of sources, including PDFs and Excel files. Users can interact with text-based queries as well as voice commands. To make this simple, we integrate Whisper AI for speech recognition, allowing users to ask questions verbally, but Google's text speech (GTTS) gives the answer generated by AI to speak the spoken language. Convert to feedback. An important feature of our system is the ability to store and show information at a granular level. Instead of dealing with the entire document, organize and retrieve relevant sections to ensure more detailed answers. FAISS-based similarity search helps you efficiently find the most relevant information in large data records, but Mistral AI produces documents to improve the quality and consistency o f answers will be improved. User can perform a profound search process, extract meaningful knowledge, and interact in a seamless, intuitive way using AI-controlled knowledge .Ultimately,ourrag system bridges the gap between data calls andAI- controlled content generation, making information accessible and easy implementable in a variety of applications. Keywords: Large Language Models (LLMs), Data Pipelines, Data Retrieval
39
Meng, Wenjun, Yuzhe Li, Lili Chen et Zhaomin Dong. « Using the Retrieval-Augmented Generation to Improve the Question-Answering System in Human Health Risk Assessment : The Development and Application ». Electronics 14, no 2 (20 janvier 2025) : 386. https://doi.org/10.3390/electronics14020386.
Texte intégralRésumé :
While large language models (LLMs) are vital for retrieving relevant information from extensive knowledge bases, they always face challenges, including high costs and issues of credibility. Here, we developed a question answering system focused on human health risk using Retrieval-Augmented Generation (RAG). We first proposed a framework to generate question–answer pairs, resulting in 300 high-quality pairs across six subfields. Subsequently, we created both a Naive RAG and an Advanced RAG-based Question-Answering (Q&A) system. Performance evaluation of the 300 question–answer pairs in individual research subfields demonstrated that the Advanced RAG outperformed traditional LLMs (including ChatGPT and ChatGLM) and Naive RAG. Finally, we integrated the developed module for a single subfield to launch a multi-knowledge base question answering system. Our study represents a novel application of RAG technology and LLMs to optimize knowledge retrieval methods in human health risk assessment.
40
Liang, Xujian, et Zhaoquan Gu. « Fast Think-on-Graph : Wider, Deeper and Faster Reasoning of Large Language Model on Knowledge Graph ». Proceedings of the AAAI Conference on Artificial Intelligence 39, no 23 (11 avril 2025) : 24558–66. https://doi.org/10.1609/aaai.v39i23.34635.
Texte intégralRésumé :
Graph Retrieval Augmented Generation (GRAG) is a novel paradigm that takes the naive RAG system a step further by integrating graph information, such as knowledge graph (KGs), into large-scale language models (LLMs) to mitigate hallucination. However, existing GRAG still encounter limitations: 1) simple paradigms usually fail with the complex problems due to the narrow and shallow correlations capture from KGs 2) methods of strong coupling with KGs tend to be high computation cost and time consuming if the graph is dense. In this paper, we propose the Fast Think-on-Graph (FastToG), an innovative paradigm for enabling LLMs to think ``community by community" within KGs. To do this, FastToG employs community detection for deeper correlation capture and two stages community pruning - coarse and fine pruning for faster retrieval. Furthermore, we also develop two Community-to-Text methods to convert the graph structure of communities into textual form for better understanding by LLMs. Experimental results demonstrate the effectiveness of FastToG, showcasing higher accuracy, faster reasoning, and better explainability compared to the previous works.
41
Tang, Xiaqiang, Jian Li, Nan Du et Sihong Xie. « Adapting to Non-Stationary Environments : Multi-Armed Bandit Enhanced Retrieval-Augmented Generation on Knowledge Graphs ». Proceedings of the AAAI Conference on Artificial Intelligence 39, no 12 (11 avril 2025) : 12658–66. https://doi.org/10.1609/aaai.v39i12.33380.
Texte intégralRésumé :
Despite the superior performance of Large language models on many NLP tasks, they still face significant limitations in memorizing extensive world knowledge. Recent studies have demonstrated that leveraging the Retrieval-Augmented Generation (RAG) framework, combined with Knowledge Graphs that encapsulate extensive factual data in a structured format, robustly enhances the reasoning capabilities of LLMs. However, deploying such systems in real-world scenarios presents challenges: the continuous evolution of non-stationary environments may lead to performance degradation and user satisfaction requires a careful balance of performance and responsiveness. To address these challenges, we introduce a Multi-objective Multi-Armed Bandit enhanced RAG framework, supported by multiple retrieval methods with diverse capabilities under rich and evolving retrieval contexts in practice. Within this framework, each retrieval method is treated as a distinct "arm''. The system utilizes real-time user feedback to adapt to dynamic environments, by selecting the appropriate retrieval method based on input queries and the historical multi-objective performance of each arm. Extensive experiments conducted on two benchmark KGQA datasets demonstrate that our method significantly outperforms baseline methods in non-stationary settings while achieving state-of-the-art performance in station environments.
42
Yu, Runjie, Weizhou Huang, Shuhan Bai, Jian Zhou et Fei Wu. « AquaPipe : A Quality-Aware Pipeline for Knowledge Retrieval and Large Language Models ». Proceedings of the ACM on Management of Data 3, no 1 (10 février 2025) : 1–26. https://doi.org/10.1145/3709661.
Texte intégralRésumé :
The knowledge retrieval methods such as Approximate Nearest Neighbor Search (ANNS) significantly enhance the generation quality of Large Language Models (LLMs) by introducing external knowledge, and this method is called Retrieval-augmented generation (RAG). However, due to the rapid growth of data size, ANNS tends to store large-scale data on disk, which greatly increases the response time of RAG systems. This paper presents AquaPipe, which pipelines the execution of disk-based ANNS and the LLM prefill phase in an RAG system, effectively overlapping the latency of knowledge retrieval and model inference to enhance the overall performance, while guaranteeing data quality. First, ANNS's recall-aware prefetching strategy enables the early return of partial text with acceptable accuracy so the prefill phase can launch before getting the full results. Then, we adaptively choose the remove-after-prefill or re-prefill strategies based on the LLM cost model to effectively correct disturbed pipelines caused by wrong early returns. Finally, the pipelined prefill dynamically changes the granularity of chunk size to balance the overlap efficiency and GPU efficiency, adjusting to ANNS tasks that converge at different speeds. Our experiments have demonstrated the effectiveness of AquaPipe. It successfully masks the latency of disk-based ANNS by 56% to 99%, resulting in a 1.3× to 2.6× reduction of the response time of the RAG, while the extra recall loss caused by prefetching is limited to approximately 1%.
43
Hao, Jin-Xing, Lei Chen et Luyao Meng. « Advancing Large Language Models with Enhanced Retrieval-Augmented Generation : Evidence from Biological UAV Swarm Control ». Drones 9, no 5 (10 mai 2025) : 361. https://doi.org/10.3390/drones9050361.
Texte intégralRésumé :
As research on biological unmanned aerial vehicle (UAV) swarm control has blossomed, professionals face increasing time and cognitive pressure in mastering the rapidly growing domain knowledge. Although recent general large language models (LLMs) may augment human cognitive capabilities, they still face significant hallucination and interpretability issues in domain-specific applications. To address these challenges, this study designs and evaluates a domain-specific LLM for the biological UAV swarm control using an enhanced Retrieval-Augmented Generation (RAG) framework. In particular, this study proposes an element-based chunking strategy to build the domain-specific knowledge base and develops novel hybrid retrieval and reranking modules to improve the classical RAG framework. This study also carefully conducts automatic and expert evaluations of our domain-specific LLM, demonstrating the advantages of our model regarding accuracy, relevance, and human alignment.
44
Lakatos, Róbert, Péter Pollner, András Hajdu et Tamás Joó. « Investigating the Performance of Retrieval-Augmented Generation and Domain-Specific Fine-Tuning for the Development of AI-Driven Knowledge-Based Systems ». Machine Learning and Knowledge Extraction 7, no 1 (10 février 2025) : 15. https://doi.org/10.3390/make7010015.
Texte intégralRésumé :
Generative large language models (LLMs) have revolutionized the development of knowledge-based systems, enabling new possibilities in applications like ChatGPT, Bing, and Gemini. Two key strategies for domain adaptation in these systems are Domain-Specific Fine-Tuning (DFT) and Retrieval-Augmented Generation (RAG). In this study, we evaluate the performance of RAG and DFT on several LLM architectures, including GPT-J-6B, OPT-6.7B, LLaMA, and LLaMA-2. We use the ROUGE, BLEU, and METEOR scores to evaluate the performance of the models. We also measure the performance of the models with our own designed cosine similarity-based Coverage Score (CS). Our results, based on experiments across multiple datasets, show that RAG-based systems consistently outperform those fine-tuned with DFT. Specifically, RAG models outperform DFT by an average of 17% in ROUGE, 13% in BLEU, and 36% in CS. At the same time, DFT achieves only a modest advantage in METEOR, suggesting slightly better creative capabilities. We also highlight the challenges of integrating RAG with DFT, as such integration can lead to performance degradation. Furthermore, we propose a simplified RAG-based architecture that maximizes efficiency and reduces hallucination, underscoring the advantages of RAG in building reliable, domain-adapted knowledge systems.
45
Alam, Sirojul, Jaka Abdul Jabar, Fauzi Abdurrachman, Bambang Suharjo et H. A. Danang Rimbawa. « Improving Large Language Model’s Ability to Find the Words Relationship ». Jurnal Bumigora Information Technology (BITe) 6, no 2 (9 novembre 2024) : 141–48. https://doi.org/10.30812/bite.v6i2.4127.
Texte intégralRésumé :
Background: It is still possible to enhance the capabilities of popular and widely used large language models (LLMs) such as Generative Pre-trained Transformer (GPT). Using the Retrieval-Augmented Generation (RAG) architecture is one method of achieving enhancement. This architectural approach incorporates outside data into the model to improve LLM capabilities. Objective: The aim of this research is to prove that the RAG can help LLMs respond with greater precision and rationale. Method: The method used in this work is utilizing Huggingface Application Programming Interface (API) for word embedding, store and find the relationship of the words. Result: The results show how well RAG performs, as the attractively rendered graph makes clear. The knowledge that has been obtained is logical and understandable, such as the word Logistic Regression that related to accuracy, F1 score, and defined as a simple and the best model compared to Naïve Bayes and Support Vector Machine (SVM) model. Conclusion: The conclusion is RAG helps LLMs to improve its capability well.
46
Ivanov, Oleg Anatolievich, Aleksey Andreevich Romanov, Anton Alexandrovich Titorenko et Irina Vladimirovna Rodicheva. « AI ASSISTANT FOR THE ADAPTATION OF IT SPECIALISTS IN THE ITS FIELD : ISSUES OF RELEVANCE, ARCHITECTURE, SECURITY, AND ETHICS IN THE CONTEXT OF DIGITALIZATION IN THE TRANSPORT SECTOR ». World of transport and technological machines 88, no 1-1 (2025) : 111–20. https://doi.org/10.33979/2073-7432-2025-1-1(88)-111-120.
Texte intégralRésumé :
The article discusses the use of a personalized AI assistant based on RAG (Retrieval-Augmented Generation) for the adaptation of IT specialists in the ITS field. The capabilities of the technology, the solution architecture, security and ethics issues are discussed. The use of RAG on domain knowledge sources for the adaptation of employees in the transport industry contributes to digital transformation, advanced training of personnel and sustainable development of ITS in the regions.
47
Kim, Junseo, Seok Jun Kim, Junseok Ahn et Suehyun Lee. « LLM-Based Response Generation for Korean Adolescents : A Study Using the NAVER Knowledge iN Q&A Dataset with RAG ». Healthcare Informatics Research 31, no 2 (30 avril 2025) : 136–45. https://doi.org/10.4258/hir.2025.31.2.136.
Texte intégralRésumé :
Objectives: This research aimed to develop a retrieval-augmented generation (RAG) based large language model (LLM) system that offers personalized and reliable responses to a wide range of concerns raised by Korean adolescents. Our work focuses on building a culturally reflective dataset and on designing and validating the system’s effectiveness by comparing the answer quality of RAG-based models with non-RAG models.Methods: Data were collected from the NAVER Knowledge iN platform, concentrating on posts that featured adolescents’ questions and corresponding expert responses during the period 2014–2024. The dataset comprises 3,874 cases, categorized by key negative emotions and the primary sources of worry. The data were processed to remove irrelevant or redundant content and then classified into general and detailed causes. The RAG-based model employed FAISS for similarity-based retrieval of the top three reference cases and used GPT-4o mini for response generation. The responses generated with and without RAG were evaluated using several metrics.Results: RAG-based responses outperformed non-RAG responses across all evaluation metrics. Key findings indicate that RAG-based responses delivered more specific, empathetic, and actionable guidance, particularly when addressing complex emotional and situational concerns. The analysis revealed that family relationships, peer interactions, and academic stress are significant factors affecting adolescents’ worries, with depression and stress frequently co-occurring.Conclusions: This study demonstrates the potential of RAG-based LLMs to address the diverse and culture-specific worries of Korean adolescents. By integrating external knowledge and offering personalized support, the proposed system provides a scalable approach to enhancing mental health interventions for adolescents. Future research should concentrate on expanding the dataset and improving multiturn conversational capabilities to deliver even more comprehensive support.
48
Ismail, Rahmat Kurnia, Farid Widyatama, Ilham Mirwansyah Wibawa, Zilmas Arjuna Brata, Ukasyah, Ghitha Afina Nelistiani et Howon Kim. « Enhancing Security Operations Center : Wazuh Security Event Response with Retrieval-Augmented-Generation-Driven Copilot ». Sensors 25, no 3 (31 janvier 2025) : 870. https://doi.org/10.3390/s25030870.
Texte intégralRésumé :
The sophistication of cyberthreats demands more efficient and intelligent tools to support Security Operations Centers (SOCs) in managing and mitigating incidents. To address this, we developed the Security Event Response Copilot (SERC), a system designed to assist analysts in responding to and mitigating security breaches more effectively. SERC integrates two core components: (1) security event data extraction using Retrieval-Augmented Generation (RAG) methods, and (2) LLM-based incident response guidance. This paper specifically utilizes Wazuh, an open-source Security Information and Event Management (SIEM) platform, as the foundation for capturing, analyzing, and correlating security events from endpoints. SERC leverages Wazuh’s capabilities to collect real-time event data and applies a RAG approach to retrieve context-specific insights from three vectorized data collections: incident response knowledge, the MITRE ATT&CK framework, and the NIST Cybersecurity Framework (CSF) 2.0. This integration bridges strategic risk management and tactical intelligence, enabling precise identification of adversarial tactics and techniques while adhering to best practices in cybersecurity. The results demonstrate the potential of combining structured threat intelligence frameworks with AI-driven models, empowered by Wazuh’s robust SIEM capabilities, to address the dynamic challenges faced by SOCs in today’s complex cybersecurity environment.
49
Posedaru, Bogdan-Stefan, Florin-Valeriu Pantelimon, Mihai-Nicolae Dulgheru et Tiberiu-Marian Georgescu. « Artificial Intelligence Text Processing Using Retrieval-Augmented Generation : Applications in Business and Education Fields ». Proceedings of the International Conference on Business Excellence 18, no 1 (1 juin 2024) : 209–22. http://dx.doi.org/10.2478/picbe-2024-0018.
Texte intégralRésumé :
Abstract The article studies the current text processing tools based on Artificial Intelligence. A literature review is done emphasizing the dynamic evolution of AI-powered text analytics, having as its central tool ChatGPT and its capabilities. The focus is centered on the techniques and methods that are using embeddings in order to improve large language models (LLMs). In this paper is analyzed the current situation of the literature in terms of text processing using Retrieval-Augmented Generation and is highlighted the potential of this technology to enhance the interpretability and trust in applications critical, such as those related to education or business. AI has revolutionized natural language processing (NLP), which facilitated the machines to interpret and generate text efficiently and accurately. In addition, large language models with external knowledge bases have been developed. These are used to produce more accurate and contextually relevant text responses. This approach is called Retrieval-Augmented Generation (RAG is one of the most significant recent advancements in this field. Based on our study, two use cases are implemented to show the applicability of our study: one related to education and one related to business IT-related documents. The methodology describes the techniques used. This includes retrieval-augmented generation and embedding stored using vector databases. Our custom models are evaluated by comparing them to the general ones, without embeddings, showing superior performance. The article highlights remarkable progress in Retrieval-Augmented Generation (RAG), which is used for AI text processing with a focus on business and education fields. Further in this paper, many of the most significant highlights are presented, which include a scalable framework for AI applications, a new integration of Retrieval-Augmented Generation and embeddings, practical application demonstrations, bridging gaps in the analysis op AI text, significant development in AI performance and optimizing educational and business processes.
50
Xiong, Jie, Lingmin Pan, Yanjiao Liu, Lei Zhu, Lizhuo Zhang et Siqiao Tan. « Enhancing Plant Protection Knowledge with Large Language Models : A Fine-Tuned Question-Answering System Using LoRA ». Applied Sciences 15, no 7 (1 avril 2025) : 3850. https://doi.org/10.3390/app15073850.
Texte intégralRésumé :
To enhance the accessibility and accuracy of plant protection knowledge for agricultural practitioners, this study develops an intelligent question-answering (QA) system based on a large language model (LLM). A local knowledge base was constructed by vectorizing 7000 research papers and books in the field of plant protection, from which 568 representative papers were selected to generate QA data using an LLM. After data cleaning and filtering, a fine-tuning dataset comprising 9000 question–answer pairs was curated. To optimize the model’s performance, low-rank adaptation (LoRA) was applied to the InterLM-20B model, resulting in the InterLM-20B-LoRA, which was integrated with Langchain-ChatChat and the local knowledge base to develop the QA system. Additionally, retrieval-augmented generation (RAG) technology was implemented to enhance response accuracy by enabling the model to retrieve relevant field-specific knowledge before generating answers, effectively mitigating the risk of hallucinated information. The experimental results demonstrate that the proposed system achieves an answer accuracy of 97%, highlighting its potential as an advanced solution for intelligent agricultural QA services.