{"id":{"repo_id":"vt","oai_identifier":"oai:vtechworks.lib.vt.edu:10919/135522"},"canonical_url":"https://search.dev.ndltd.org/etd/vt/oai:vtechworks.lib.vt.edu:10919/135522","repository":{"repo_id":"vt","name":"Virginia Tech","base_url":"https://vtechworks.lib.vt.edu/oai/request"},"display":{"title":"QuOTE: Question-Oriented Text Embeddings","abstract":"We present QuOTE (Question-Oriented Text Embeddings), a novel enhancement to retrieval- augmented generation (RAG) systems, aimed at improving document representation for accurate and nuanced retrieval. Unlike traditional RAG pipelines, which rely on embed- ding raw text chunks, QuOTE augments chunks with hypothetical questions that the chunk can potentially answer, enriching the representation space. This better aligns document embeddings with user query semantics, and helps address issues such as ambiguity and context-dependent relevance. Through extensive experiments across diverse benchmarks, we demonstrate that QuOTE significantly enhances retrieval accuracy, including in multi-hop question-answering tasks. Our findings highlight the versatility of question generation as a fundamental indexing strategy, opening new avenues for integrating question generation into retrieval-based AI pipelines.","abstract_html":"We present QuOTE (Question-Oriented Text Embeddings), a novel enhancement to retrieval- augmented generation (RAG) systems, aimed at improving document representation for accurate and nuanced retrieval. Unlike traditional RAG pipelines, which rely on embed- ding raw text chunks, QuOTE augments chunks with hypothetical questions that the chunk can potentially answer, enriching the representation space. This better aligns document embeddings with user query semantics, and helps address issues such as ambiguity and context-dependent relevance. Through extensive experiments across diverse benchmarks, we demonstrate that QuOTE significantly enhances retrieval accuracy, including in multi-hop question-answering tasks. Our findings highlight the versatility of question generation as a fundamental indexing strategy, opening new avenues for integrating question generation into retrieval-based AI pipelines.","abstract_has_math":false,"creators":["Neeser, Andrew Kyle"],"institution":"Virginia Tech","degree_name":"Master of Science","degree_level":"masters","degree_discipline":"Computer Science & Applications","degree_department":"Computer Science and#38; Applications","school":null,"contributors":[],"advisors":[],"committee_chairs":["Ramakrishnan, Narendran"],"committee_members":["Latimer, Chris","Lu, Chang Tien"],"year":2025,"date_issued":"2025-06-13","date_published":"2025-06-13","updated_at":"2026-07-22T22:20:00Z","subjects":["Retrieval-Augmented Generation","Synthetic Question Generation","Document Representation","Information Retrieval"],"languages":["en"],"rights":["In Copyright"],"rights_urls":["http://rightsstatements.org/vocab/InC/1.0/"],"identifier_entries":[{"key":"dc:identifier.other","label":"Dc Identifier Other","values":["vt_gsexam:44058"],"render_values":[{"text":"vt_gsexam:44058","href":null,"code":true}]}]},"links":{"outbound_url":"https://hdl.handle.net/10919/135522","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.committeechair","label":"Committee Chair","values":["Ramakrishnan, Narendran"]},{"key":"dc:contributor.committeemember","label":"Committee Member","values":["Latimer, Chris","Lu, Chang Tien"]},{"key":"dc:contributor.department","label":"Department","values":["Computer Science and#38; Applications"]},{"key":"dc:creator","label":"Author","values":["Neeser, Andrew Kyle"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2025-06-14T08:02:08Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2025-06-14T08:02:08Z"]},{"key":"dc:date.issued","label":"Date","values":["2025-06-13"]},{"key":"dc:publisher","label":"Institution","values":["Virginia Tech"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science & Applications"]},{"key":"thesis:degree_level","label":"Degree Level","values":["masters"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["Virginia Polytechnic Institute and State University"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Retrieval-Augmented Generation","Synthetic Question Generation","Document Representation","Information Retrieval"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["In Copyright"]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://rightsstatements.org/vocab/InC/1.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.other","label":"Dc Identifier Other","values":["vt_gsexam:44058"]},{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/10919/135522"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["We present QuOTE (Question-Oriented Text Embeddings), a novel enhancement to retrieval- augmented generation (RAG) systems, aimed at improving document representation for accurate and nuanced retrieval. Unlike traditional RAG pipelines, which rely on embed- ding raw text chunks, QuOTE augments chunks with hypothetical questions that the chunk can potentially answer, enriching the representation space. This better aligns document embeddings with user query semantics, and helps address issues such as ambiguity and context-dependent relevance. Through extensive experiments across diverse benchmarks, we demonstrate that QuOTE significantly enhances retrieval accuracy, including in multi-hop question-answering tasks. Our findings highlight the versatility of question generation as a fundamental indexing strategy, opening new avenues for integrating question generation into retrieval-based AI pipelines."]},{"key":"dc:description.abstractgeneral","label":"General Abstract","values":["Modern artificial intelligence tools often help users by searching through large collections of documents and then using those search results to generate answers. This process can sometimes misinterpret a question or miss important connections in the text. In our work, we introduce QuOTE, a simple yet powerful method that teaches the system to think in terms of questions: each piece of text is paired with relevant, hypothetical questions it could answer. By organizing information around questions and answers, QuOTE creates clearer, more meaningful representations of documents. In tests that include cases where answers require combining information from different parts of a document, QuOTE consistently retrieves more accurate and relevant information than traditional approaches. This question-based indexing approach makes search-and-answer systems more reliable and could enhance a wide range of everyday tools, from virtual assistants to online help desks."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["Master of Science"]},{"key":"dc:format.medium","label":"Dc Format Medium","values":["ETD"]},{"key":"dc:title","label":"Title","values":["QuOTE: Question-Oriented Text Embeddings"]}]}],"canonical_facts":{"dc:contributor.committeechair":["Ramakrishnan, Narendran"],"dc:contributor.committeemember":["Latimer, Chris","Lu, Chang Tien"],"dc:contributor.department":["Computer Science and#38; Applications"],"dc:creator":["Neeser, Andrew Kyle"],"dc:date.accessioned":["2025-06-14T08:02:08Z"],"dc:date.available":["2025-06-14T08:02:08Z"],"dc:date.issued":["2025-06-13"],"dc:description.abstract":["We present QuOTE (Question-Oriented Text Embeddings), a novel enhancement to retrieval- augmented generation (RAG) systems, aimed at improving document representation for accurate and nuanced retrieval. Unlike traditional RAG pipelines, which rely on embed- ding raw text chunks, QuOTE augments chunks with hypothetical questions that the chunk can potentially answer, enriching the representation space. This better aligns document embeddings with user query semantics, and helps address issues such as ambiguity and context-dependent relevance. Through extensive experiments across diverse benchmarks, we demonstrate that QuOTE significantly enhances retrieval accuracy, including in multi-hop question-answering tasks. Our findings highlight the versatility of question generation as a fundamental indexing strategy, opening new avenues for integrating question generation into retrieval-based AI pipelines."],"dc:description.abstractgeneral":["Modern artificial intelligence tools often help users by searching through large collections of documents and then using those search results to generate answers. This process can sometimes misinterpret a question or miss important connections in the text. In our work, we introduce QuOTE, a simple yet powerful method that teaches the system to think in terms of questions: each piece of text is paired with relevant, hypothetical questions it could answer. By organizing information around questions and answers, QuOTE creates clearer, more meaningful representations of documents. In tests that include cases where answers require combining information from different parts of a document, QuOTE consistently retrieves more accurate and relevant information than traditional approaches. This question-based indexing approach makes search-and-answer systems more reliable and could enhance a wide range of everyday tools, from virtual assistants to online help desks."],"dc:description.degree":["Master of Science"],"dc:format.medium":["ETD"],"dc:identifier.other":["vt_gsexam:44058"],"dc:identifier.uri":["https://hdl.handle.net/10919/135522"],"dc:language.iso":["en"],"dc:publisher":["Virginia Tech"],"dc:rights":["In Copyright"],"dc:rights.uri":["http://rightsstatements.org/vocab/InC/1.0/"],"dc:subject":["Retrieval-Augmented Generation","Synthetic Question Generation","Document Representation","Information Retrieval"],"dc:title":["QuOTE: Question-Oriented Text Embeddings"],"dc:type":["Thesis"],"thesis:degree_discipline":["Computer Science & Applications"],"thesis:degree_level":["masters"],"thesis:degree_name":["Master of Science"],"thesis:institution_name":["Virginia Polytechnic Institute and State University"]},"updated_at":"2026-07-22T22:20:00Z"}