{"id":{"repo_id":"queens","oai_identifier":"oai:queensu.scholaris.ca:1974/36013"},"canonical_url":"https://search.dev.ndltd.org/etd/queens/oai:queensu.scholaris.ca:1974/36013","repository":{"repo_id":"queens","name":"Queens University","base_url":"https://qspace.library.queensu.ca/server/oai/request"},"display":{"title":"AI-Serve: Empowering Service Provisioning with Conversational and Generative AI","abstract":"The global conversational AI market has been growing rapidly with widely adopted applications in diverse domains, including customer service, finance, education, and healthcare. Voice-enabled assistants such as Siri, Alexa, and Google Assistant are used by hundreds of millions of people worldwide nowadays, reflecting a strong and growing demand for Artificial Intelligence (AI)-powered conversational service systems. However, these systems still face significant limitations as assistive agents, particularly when reliable and personalized support is required in sensitive contexts. They also demonstrate limited analytical capability when dealing with unstructured, domain-specific information, including clinical notes, patient queries, and doctor patient conversations, where accurate interpretation is critical. Recent advances in Large Language Models (LLMs) have greatly improved the ability of conversational systems to handle unstructured text and reasoning tasks. However, these models also suffer from drawbacks such as hallucination, inconsistent reliability, and the lack of well-defined metrics to evaluate their responses, which limits the safe deployment in sensitive domains such as healthcare. This dissertation contributes to advancing conversational AI in three key aspects: (a) enhancing assistive capabilities through a voice-enabled chatbot framework designed to support older adults’ daily care needs; (b) strengthening query services by integrating Natural Language Processing and Machine Learning modules to interpret complex medical data as demonstrated through two case scenarios: Osteoarthritis (OA) pain severity detection and Dementia and Alzheimer's Disease (AD) classification; and (c) improving the reliability of generative AI with a Multi-Agent Summarization and Auto Evaluation (MASA) that systematically generates medical text summaries and evaluates their quality and factual consistency. Together, these contributions bridge the gap between general-purpose conversational systems and specialized real-world, domain-sensitive conversational assistive and knowledge services.","abstract_html":"The global conversational AI market has been growing rapidly with widely adopted applications in diverse domains, including customer service, finance, education, and healthcare. Voice-enabled assistants such as Siri, Alexa, and Google Assistant are used by hundreds of millions of people worldwide nowadays, reflecting a strong and growing demand for Artificial Intelligence (AI)-powered conversational service systems. However, these systems still face significant limitations as assistive agents, particularly when reliable and personalized support is required in sensitive contexts. They also demonstrate limited analytical capability when dealing with unstructured, domain-specific information, including clinical notes, patient queries, and doctor patient conversations, where accurate interpretation is critical. Recent advances in Large Language Models (LLMs) have greatly improved the ability of conversational systems to handle unstructured text and reasoning tasks. However, these models also suffer from drawbacks such as hallucination, inconsistent reliability, and the lack of well-defined metrics to evaluate their responses, which limits the safe deployment in sensitive domains such as healthcare. This dissertation contributes to advancing conversational AI in three key aspects: (a) enhancing assistive capabilities through a voice-enabled chatbot framework designed to support older adults’ daily care needs; (b) strengthening query services by integrating Natural Language Processing and Machine Learning modules to interpret complex medical data as demonstrated through two case scenarios: Osteoarthritis (OA) pain severity detection and Dementia and Alzheimer's Disease (AD) classification; and (c) improving the reliability of generative AI with a Multi-Agent Summarization and Auto Evaluation (MASA) that systematically generates medical text summaries and evaluates their quality and factual consistency. Together, these contributions bridge the gap between general-purpose conversational systems and specialized real-world, domain-sensitive conversational assistive and knowledge services.","abstract_has_math":false,"creators":["Chen, Yuhao"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":"Computing","school":null,"contributors":[],"advisors":["Zulkernine, Farhana"],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026-01-22","date_published":"2026-01-22","updated_at":"2026-07-27T20:35:39Z","subjects":["Conversational Artificial Intelligence","Service-Oriented AI Systems","Healthcare AI","Medical Text Analytics","Electronic Medical Records (EMR) Analysis","Personalized Conversational Agents","Analytical Question Answering","LLMs-as-Judges","Multi-Agent Summarization","Hallucination Detection","Expert–LLM Agreement Analysis"],"languages":["eng"],"rights":["Attribution-NonCommercial-NoDerivatives 4.0 International"],"rights_urls":["http://creativecommons.org/licenses/by-nc-nd/4.0/"],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/1974/36013","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.department","label":"Department","values":["Computing"]},{"key":"dc:contributor.supervisor","label":"Supervisor","values":["Zulkernine, Farhana"]},{"key":"dc:creator","label":"Author","values":["Chen, Yuhao"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-01-22T15:25:35Z"]},{"key":"dc:date.issued","label":"Date","values":["2026-01-22"]},{"key":"dc:type","label":"Dc Type","values":["thesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Conversational Artificial Intelligence","Service-Oriented AI Systems","Healthcare AI","Medical Text Analytics","Electronic Medical Records (EMR) Analysis","Personalized Conversational Agents","Analytical Question Answering","LLMs-as-Judges","Multi-Agent Summarization","Hallucination Detection","Expert–LLM Agreement Analysis"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Attribution-NonCommercial-NoDerivatives 4.0 International"]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://creativecommons.org/licenses/by-nc-nd/4.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/1974/36013"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["The global conversational AI market has been growing rapidly with widely adopted applications in diverse domains, including customer service, finance, education, and healthcare. Voice-enabled assistants such as Siri, Alexa, and Google Assistant are used by hundreds of millions of people worldwide nowadays, reflecting a strong and growing demand for Artificial Intelligence (AI)-powered conversational service systems. However, these systems still face significant limitations as assistive agents, particularly when reliable and personalized support is required in sensitive contexts. They also demonstrate limited analytical capability when dealing with unstructured, domain-specific information, including clinical notes, patient queries, and doctor patient conversations, where accurate interpretation is critical. Recent advances in Large Language Models (LLMs) have greatly improved the ability of conversational systems to handle unstructured text and reasoning tasks. However, these models also suffer from drawbacks such as hallucination, inconsistent reliability, and the lack of well-defined metrics to evaluate their responses, which limits the safe deployment in sensitive domains such as healthcare. This dissertation contributes to advancing conversational AI in three key aspects: (a) enhancing assistive capabilities through a voice-enabled chatbot framework designed to support older adults’ daily care needs; (b) strengthening query services by integrating Natural Language Processing and Machine Learning modules to interpret complex medical data as demonstrated through two case scenarios: Osteoarthritis (OA) pain severity detection and Dementia and Alzheimer's Disease (AD) classification; and (c) improving the reliability of generative AI with a Multi-Agent Summarization and Auto Evaluation (MASA) that systematically generates medical text summaries and evaluates their quality and factual consistency. Together, these contributions bridge the gap between general-purpose conversational systems and specialized real-world, domain-sensitive conversational assistive and knowledge services."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["PhD"]},{"key":"dc:title","label":"Title","values":["AI-Serve: Empowering Service Provisioning with Conversational and Generative AI"]}]}],"canonical_facts":{"dc:contributor.department":["Computing"],"dc:contributor.supervisor":["Zulkernine, Farhana"],"dc:creator":["Chen, Yuhao"],"dc:date.accessioned":["2026-01-22T15:25:35Z"],"dc:date.issued":["2026-01-22"],"dc:description.abstract":["The global conversational AI market has been growing rapidly with widely adopted applications in diverse domains, including customer service, finance, education, and healthcare. Voice-enabled assistants such as Siri, Alexa, and Google Assistant are used by hundreds of millions of people worldwide nowadays, reflecting a strong and growing demand for Artificial Intelligence (AI)-powered conversational service systems. However, these systems still face significant limitations as assistive agents, particularly when reliable and personalized support is required in sensitive contexts. They also demonstrate limited analytical capability when dealing with unstructured, domain-specific information, including clinical notes, patient queries, and doctor patient conversations, where accurate interpretation is critical. Recent advances in Large Language Models (LLMs) have greatly improved the ability of conversational systems to handle unstructured text and reasoning tasks. However, these models also suffer from drawbacks such as hallucination, inconsistent reliability, and the lack of well-defined metrics to evaluate their responses, which limits the safe deployment in sensitive domains such as healthcare. This dissertation contributes to advancing conversational AI in three key aspects: (a) enhancing assistive capabilities through a voice-enabled chatbot framework designed to support older adults’ daily care needs; (b) strengthening query services by integrating Natural Language Processing and Machine Learning modules to interpret complex medical data as demonstrated through two case scenarios: Osteoarthritis (OA) pain severity detection and Dementia and Alzheimer's Disease (AD) classification; and (c) improving the reliability of generative AI with a Multi-Agent Summarization and Auto Evaluation (MASA) that systematically generates medical text summaries and evaluates their quality and factual consistency. Together, these contributions bridge the gap between general-purpose conversational systems and specialized real-world, domain-sensitive conversational assistive and knowledge services."],"dc:description.degree":["PhD"],"dc:identifier.uri":["https://hdl.handle.net/1974/36013"],"dc:language.iso":["eng"],"dc:rights":["Attribution-NonCommercial-NoDerivatives 4.0 International"],"dc:rights.uri":["http://creativecommons.org/licenses/by-nc-nd/4.0/"],"dc:subject":["Conversational Artificial Intelligence","Service-Oriented AI Systems","Healthcare AI","Medical Text Analytics","Electronic Medical Records (EMR) Analysis","Personalized Conversational Agents","Analytical Question Answering","LLMs-as-Judges","Multi-Agent Summarization","Hallucination Detection","Expert–LLM Agreement Analysis"],"dc:title":["AI-Serve: Empowering Service Provisioning with Conversational and Generative AI"],"dc:type":["thesis"]},"updated_at":"2026-07-27T20:35:39Z"}