{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/102460"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/102460","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Textual entailment from image caption denotations","abstract":"Understanding the meaning of linguistic expressions is a fundamental task of natural language processing. While distributed representations have become a powerful technique for modeling lexical semantics, but they have traditionally relied on ungrounded text corpora to identify semantically similar words. In contrast, this thesis explicitly models the denotation of linguistic expressions by building representations from grounded image captions. This allows us to use descriptions of the world to learn connections that would be difficult to identify in text-based corpora. In particular, we explore novel approaches to entailment that capture everyday world knowledge missing from other NLP tasks, on both existing datasets and our own new dataset. We also present a novel embedding model that produces phrase representations that are informed by our grounded representation. We conclude with an analysis of how grounded embeddings differ from standard distributional embeddings and suggestions for future refinement of this approach.","abstract_html":"Understanding the meaning of linguistic expressions is a fundamental task of natural language processing. While distributed representations have become a powerful technique for modeling lexical semantics, but they have traditionally relied on ungrounded text corpora to identify semantically similar words. In contrast, this thesis explicitly models the denotation of linguistic expressions by building representations from grounded image captions. This allows us to use descriptions of the world to learn connections that would be difficult to identify in text-based corpora. In particular, we explore novel approaches to entailment that capture everyday world knowledge missing from other NLP tasks, on both existing datasets and our own new dataset. We also present a novel embedding model that produces phrase representations that are informed by our grounded representation. We conclude with an analysis of how grounded embeddings differ from standard distributional embeddings and suggestions for future refinement of this approach.","abstract_has_math":false,"creators":["Lai, Alice Yingming"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Hockenmaier, Julia","Erk, Katrin","Roth, Dan","Zhai, ChengXiang"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2019,"date_issued":"2019-02-06T19:36:24Z","date_published":"2019-02-06T19:36:24Z","updated_at":"2026-07-22T22:24:40Z","subjects":["textual entailment","semantic representations","natural language processing"],"languages":["en"],"rights":["Copyright 2018 Alice Lai"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/102460","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Hockenmaier, Julia","Erk, Katrin","Roth, Dan","Zhai, ChengXiang"]},{"key":"dc:creator","label":"Author","values":["Lai, Alice Yingming"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2019-02-06T19:36:24Z","2018-12-03","2018-12"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph.D."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["textual entailment","semantic representations","natural language processing"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2018 Alice Lai"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/102460"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Understanding the meaning of linguistic expressions is a fundamental task of natural language processing. While distributed representations have become a powerful technique for modeling lexical semantics, but they have traditionally relied on ungrounded text corpora to identify semantically similar words. In contrast, this thesis explicitly models the denotation of linguistic expressions by building representations from grounded image captions. This allows us to use descriptions of the world to learn connections that would be difficult to identify in text-based corpora. In particular, we explore novel approaches to entailment that capture everyday world knowledge missing from other NLP tasks, on both existing datasets and our own new dataset. We also present a novel embedding model that produces phrase representations that are informed by our grounded representation. We conclude with an analysis of how grounded embeddings differ from standard distributional embeddings and suggestions for future refinement of this approach.","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2019-02-05 without embargo terms","The student, Alice Lai, accepted the attached license on 2018-11-30 at 15:38.","The student, Alice Lai, submitted this Dissertation for approval on 2018-11-30 at 15:47.","This Dissertation was approved for publication on 2018-12-03 at 09:07.","DSpace SAF Submission Ingestion Package generated from Vireo submission #13166 on 2019-02-05 at 11:13:22","Made available in DSpace on 2019-02-06T19:36:24Z (GMT). No. of bitstreams: 3 LAI-DISSERTATION-2018.pdf: 2424549 bytes, checksum: 42b9f68c5e713df63ffdda60934416e8 (MD5) LICENSE.txt: 4206 bytes, checksum: 0d084741cdd9cf8b6e970ea3ddd17feb (MD5) PROQUEST_LICENSE.txt: 4552 bytes, checksum: b24cfdf433a25061921d5b11b86ba824 (MD5) Previous issue date: 2018-12-03"]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Textual entailment from image caption denotations"]}]}],"canonical_facts":{"dc:contributor":["Hockenmaier, Julia","Erk, Katrin","Roth, Dan","Zhai, ChengXiang"],"dc:creator":["Lai, Alice Yingming"],"dc:date":["2019-02-06T19:36:24Z","2018-12-03","2018-12"],"dc:description":["Understanding the meaning of linguistic expressions is a fundamental task of natural language processing. While distributed representations have become a powerful technique for modeling lexical semantics, but they have traditionally relied on ungrounded text corpora to identify semantically similar words. In contrast, this thesis explicitly models the denotation of linguistic expressions by building representations from grounded image captions. This allows us to use descriptions of the world to learn connections that would be difficult to identify in text-based corpora. In particular, we explore novel approaches to entailment that capture everyday world knowledge missing from other NLP tasks, on both existing datasets and our own new dataset. We also present a novel embedding model that produces phrase representations that are informed by our grounded representation. We conclude with an analysis of how grounded embeddings differ from standard distributional embeddings and suggestions for future refinement of this approach.","Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2019-02-05 without embargo terms","The student, Alice Lai, accepted the attached license on 2018-11-30 at 15:38.","The student, Alice Lai, submitted this Dissertation for approval on 2018-11-30 at 15:47.","This Dissertation was approved for publication on 2018-12-03 at 09:07.","DSpace SAF Submission Ingestion Package generated from Vireo submission #13166 on 2019-02-05 at 11:13:22","Made available in DSpace on 2019-02-06T19:36:24Z (GMT). No. of bitstreams: 3 LAI-DISSERTATION-2018.pdf: 2424549 bytes, checksum: 42b9f68c5e713df63ffdda60934416e8 (MD5) LICENSE.txt: 4206 bytes, checksum: 0d084741cdd9cf8b6e970ea3ddd17feb (MD5) PROQUEST_LICENSE.txt: 4552 bytes, checksum: b24cfdf433a25061921d5b11b86ba824 (MD5) Previous issue date: 2018-12-03"],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/102460"],"dc:language":["en"],"dc:rights":["Copyright 2018 Alice Lai"],"dc:subject":["textual entailment","semantic representations","natural language processing"],"dc:title":["Textual entailment from image caption denotations"],"dc:type":["text"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:24:40Z"}