{"id":{"repo_id":"cagliari","oai_identifier":"oai:iris.unica.it:11584/442245"},"canonical_url":"https://search.dev.ndltd.org/etd/cagliari/oai:iris.unica.it:11584/442245","repository":{"repo_id":"cagliari","name":"Università di Cagliari","base_url":"https://iris.unica.it/oai/request"},"display":{"title":"Knowledge engineering for semantic understanding","abstract":"This thesis addresses the challenges of improving semantic understanding in conversational agents by combining Knowledge Graphs (KGs) and Large Language Models (LLMs) within a flexible, multi-domain knowledge plugin architecture. We explore the inherent difficulties LLMs face in interpreting plain-text user queries, as well as the limitations of generative AI, particularly its tendency toward “hallucination” when generating responses. To mitigate this, we examine the complex process of extracting and structuring knowledge from raw text to construct KGs that serve as authoritative, context-rich foundations for information retrieval. The knowledge plugin architecture developed in this work enables conversational agents to leverage both KGs and LLMs to interact accurately with reliable, domain-specific sources. Our approach includes techniques such as fine-tuning and intelligent fewshot prompting to enhance LLMs’ ability to generate accurate, context-aware queries and responses over KGs. This integration significantly advances the potential for scalable, adaptable conversational agents capable of reliable information retrieval across multiple domains. The insights and techniques outlined in this thesis mark a critical step toward creating domain-agnostic AI systems that deliver semantically precise and trustworthy information.","abstract_html":"This thesis addresses the challenges of improving semantic understanding in conversational agents by combining Knowledge Graphs (KGs) and Large Language Models (LLMs) within a flexible, multi-domain knowledge plugin architecture. We explore the inherent difficulties LLMs face in interpreting plain-text user queries, as well as the limitations of generative AI, particularly its tendency toward “hallucination” when generating responses. To mitigate this, we examine the complex process of extracting and structuring knowledge from raw text to construct KGs that serve as authoritative, context-rich foundations for information retrieval. The knowledge plugin architecture developed in this work enables conversational agents to leverage both KGs and LLMs to interact accurately with reliable, domain-specific sources. Our approach includes techniques such as fine-tuning and intelligent fewshot prompting to enhance LLMs’ ability to generate accurate, context-aware queries and responses over KGs. This integration significantly advances the potential for scalable, adaptable conversational agents capable of reliable information retrieval across multiple domains. The insights and techniques outlined in this thesis mark a critical step toward creating domain-agnostic AI systems that deliver semantically precise and trustworthy information.","abstract_has_math":false,"creators":["MELONI, ANTONELLO"],"institution":"Università degli Studi di Cagliari","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["REFORGIATO RECUPERO, DIEGO ANGELO GAETANO","IANNIZZOTTO, ANTONIO"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-02-20","date_published":"2025-02-20","updated_at":"2026-07-24T01:29:55Z","subjects":["Scholarly Data","Knowledge Graph","Virtual Assistant","User Experience","Human-Robot Interaction","Information Extraction","Natural Language Processing","Course Recommendation","Conversational Agents, Labor Market, Large Language Models, Natural Language Processing, Occupational Database","Question answering","Fine-tuning","Few-shot learning","Abstract Meaning Representation","Semantic Frames","Settore INF/01 - Informatica","Settore INFO-01/A - Informatica"],"languages":["eng"],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/11584/442245","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["REFORGIATO RECUPERO, DIEGO ANGELO GAETANO","IANNIZZOTTO, ANTONIO"]},{"key":"dc:creator","label":"Author","values":["MELONI, ANTONELLO"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2025-02-20"]},{"key":"dc:publisher","label":"Institution","values":["Università degli Studi di Cagliari"]},{"key":"dc:type","label":"Dc Type","values":["info:eu-repo/semantics/doctoralThesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Scholarly Data","Knowledge Graph","Virtual Assistant","User Experience","Human-Robot Interaction","Information Extraction","Natural Language Processing","Course Recommendation","Conversational Agents, Labor Market, Large Language Models, Natural Language Processing, Occupational Database","Question answering","Fine-tuning","Few-shot learning","Abstract Meaning Representation","Semantic Frames","Settore INF/01 - Informatica","Settore INFO-01/A - Informatica"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://hdl.handle.net/11584/442245"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["This thesis addresses the challenges of improving semantic understanding in conversational agents by combining Knowledge Graphs (KGs) and Large Language Models (LLMs) within a flexible, multi-domain knowledge plugin architecture. We explore the inherent difficulties LLMs face in interpreting plain-text user queries, as well as the limitations of generative AI, particularly its tendency toward “hallucination” when generating responses. To mitigate this, we examine the complex process of extracting and structuring knowledge from raw text to construct KGs that serve as authoritative, context-rich foundations for information retrieval. The knowledge plugin architecture developed in this work enables conversational agents to leverage both KGs and LLMs to interact accurately with reliable, domain-specific sources. Our approach includes techniques such as fine-tuning and intelligent fewshot prompting to enhance LLMs’ ability to generate accurate, context-aware queries and responses over KGs. This integration significantly advances the potential for scalable, adaptable conversational agents capable of reliable information retrieval across multiple domains. The insights and techniques outlined in this thesis mark a critical step toward creating domain-agnostic AI systems that deliver semantically precise and trustworthy information."]},{"key":"dc:title","label":"Title","values":["Knowledge engineering for semantic understanding"]}]}],"canonical_facts":{"dc:contributor":["REFORGIATO RECUPERO, DIEGO ANGELO GAETANO","IANNIZZOTTO, ANTONIO"],"dc:creator":["MELONI, ANTONELLO"],"dc:date":["2025-02-20"],"dc:description":["This thesis addresses the challenges of improving semantic understanding in conversational agents by combining Knowledge Graphs (KGs) and Large Language Models (LLMs) within a flexible, multi-domain knowledge plugin architecture. We explore the inherent difficulties LLMs face in interpreting plain-text user queries, as well as the limitations of generative AI, particularly its tendency toward “hallucination” when generating responses. To mitigate this, we examine the complex process of extracting and structuring knowledge from raw text to construct KGs that serve as authoritative, context-rich foundations for information retrieval. The knowledge plugin architecture developed in this work enables conversational agents to leverage both KGs and LLMs to interact accurately with reliable, domain-specific sources. Our approach includes techniques such as fine-tuning and intelligent fewshot prompting to enhance LLMs’ ability to generate accurate, context-aware queries and responses over KGs. This integration significantly advances the potential for scalable, adaptable conversational agents capable of reliable information retrieval across multiple domains. The insights and techniques outlined in this thesis mark a critical step toward creating domain-agnostic AI systems that deliver semantically precise and trustworthy information."],"dc:identifier":["https://hdl.handle.net/11584/442245"],"dc:language":["eng"],"dc:publisher":["Università degli Studi di Cagliari"],"dc:subject":["Scholarly Data","Knowledge Graph","Virtual Assistant","User Experience","Human-Robot Interaction","Information Extraction","Natural Language Processing","Course Recommendation","Conversational Agents, Labor Market, Large Language Models, Natural Language Processing, Occupational Database","Question answering","Fine-tuning","Few-shot learning","Abstract Meaning Representation","Semantic Frames","Settore INF/01 - Informatica","Settore INFO-01/A - Informatica"],"dc:title":["Knowledge engineering for semantic understanding"],"dc:type":["info:eu-repo/semantics/doctoralThesis"]},"updated_at":"2026-07-24T01:29:55Z"}