{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/24006"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/24006","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"An Exploration of Multimodal Document Classification Strategies","abstract":"This thesis explores multimodal document classification algorithms in a unified framework. Classification algorithms are designed to exploit both text and image information, which proliferates in modern documents. We design meta-classification schemes that combine and integrate state-of-the-art text and image feature-extractors with state-of-the-art classifiers. Meta-classifiers fuse information across modalities that differ in nature and hence have more information on hand to make decisions. This thesis also discusses strategies that exploit correlations not only within a single modality but also among modalities. Techniques that exploit correlations within a modality include image meta-feature vector combination and latent Dirichlet allocation-based image meta-feature extraction. Another technique that exploits correlations between text and image cleans image with text information. Experiments on real-world databases from Wikipedia demonstrate the benefits of metaclassification for multimodal documents.","abstract_html":"This thesis explores multimodal document classification algorithms in a unified framework. Classification algorithms are designed to exploit both text and image information, which proliferates in modern documents. We design meta-classification schemes that combine and integrate state-of-the-art text and image feature-extractors with state-of-the-art classifiers. Meta-classifiers fuse information across modalities that differ in nature and hence have more information on hand to make decisions. This thesis also discusses strategies that exploit correlations not only within a single modality but also among modalities. Techniques that exploit correlations within a modality include image meta-feature vector combination and latent Dirichlet allocation-based image meta-feature extraction. Another technique that exploits correlations between text and image cleans image with text information. Experiments on real-world databases from Wikipedia demonstrate the benefits of metaclassification for multimodal documents.","abstract_has_math":false,"creators":["Chen, Scott D."],"institution":"University of Illinois at Urbana-Champaign","degree_name":"M.S.","degree_level":"Thesis","degree_discipline":"Electrical & Computer Engr","degree_department":null,"school":null,"contributors":["Moulin, Pierre"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2011,"date_issued":"2011-05-25T14:51:49Z","date_published":"2011-05-25T14:51:49Z","updated_at":"2026-07-22T22:25:23Z","subjects":["meta-classifier","classification","multimodal","document","support vector machines"],"languages":["en"],"rights":["Copyright 2011 Scott Deeann Chen"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/24006","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Moulin, Pierre"]},{"key":"dc:creator","label":"Author","values":["Chen, Scott D."]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2011-05-25T14:51:49Z","2011-05"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical & Computer Engr"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S."]},{"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":["meta-classifier","classification","multimodal","document","support vector machines"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2011 Scott Deeann Chen"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/24006"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["This thesis explores multimodal document classification algorithms in a unified framework. Classification algorithms are designed to exploit both text and image information, which proliferates in modern documents. We design meta-classification schemes that combine and integrate state-of-the-art text and image feature-extractors with state-of-the-art classifiers. Meta-classifiers fuse information across modalities that differ in nature and hence have more information on hand to make decisions. This thesis also discusses strategies that exploit correlations not only within a single modality but also among modalities. Techniques that exploit correlations within a modality include image meta-feature vector combination and latent Dirichlet allocation-based image meta-feature extraction. Another technique that exploits correlations between text and image cleans image with text information. Experiments on real-world databases from Wikipedia demonstrate the benefits of metaclassification for multimodal documents.","Item withdrawn by Alexis Thompson (athmpsn1@illinois.edu) on 2011-04-21T21:58:29Z Item was in collections: University of Illinois Theses & Dissertations (ID: 1) No. of bitstreams: 2 Thesis.zip: 11862883 bytes, checksum: 46216d135279c32e59a6597580e78534 (MD5) Chen_Scott.pdf: 983244 bytes, checksum: a9d64c963139d763a78c2020cb845c5f (MD5)","Made available in DSpace on 2011-05-25T14:51:49Z (GMT). No. of bitstreams: 3 Chen_Scott.pdf: 983243 bytes, checksum: ea4ccde7488ad18a7fc9395ceb0abd9e (MD5) license.txt: 4059 bytes, checksum: bd0ec53acac905061ab056f7df2ff7a6 (MD5) Thesis.zip: 11862883 bytes, checksum: 46216d135279c32e59a6597580e78534 (MD5)"]},{"key":"dc:title","label":"Title","values":["An Exploration of Multimodal Document Classification Strategies"]}]}],"canonical_facts":{"dc:contributor":["Moulin, Pierre"],"dc:creator":["Chen, Scott D."],"dc:date":["2011-05-25T14:51:49Z","2011-05"],"dc:description":["This thesis explores multimodal document classification algorithms in a unified framework. Classification algorithms are designed to exploit both text and image information, which proliferates in modern documents. We design meta-classification schemes that combine and integrate state-of-the-art text and image feature-extractors with state-of-the-art classifiers. Meta-classifiers fuse information across modalities that differ in nature and hence have more information on hand to make decisions. This thesis also discusses strategies that exploit correlations not only within a single modality but also among modalities. Techniques that exploit correlations within a modality include image meta-feature vector combination and latent Dirichlet allocation-based image meta-feature extraction. Another technique that exploits correlations between text and image cleans image with text information. Experiments on real-world databases from Wikipedia demonstrate the benefits of metaclassification for multimodal documents.","Item withdrawn by Alexis Thompson (athmpsn1@illinois.edu) on 2011-04-21T21:58:29Z Item was in collections: University of Illinois Theses & Dissertations (ID: 1) No. of bitstreams: 2 Thesis.zip: 11862883 bytes, checksum: 46216d135279c32e59a6597580e78534 (MD5) Chen_Scott.pdf: 983244 bytes, checksum: a9d64c963139d763a78c2020cb845c5f (MD5)","Made available in DSpace on 2011-05-25T14:51:49Z (GMT). No. of bitstreams: 3 Chen_Scott.pdf: 983243 bytes, checksum: ea4ccde7488ad18a7fc9395ceb0abd9e (MD5) license.txt: 4059 bytes, checksum: bd0ec53acac905061ab056f7df2ff7a6 (MD5) Thesis.zip: 11862883 bytes, checksum: 46216d135279c32e59a6597580e78534 (MD5)"],"dc:identifier":["http://hdl.handle.net/2142/24006"],"dc:language":["en"],"dc:rights":["Copyright 2011 Scott Deeann Chen"],"dc:subject":["meta-classifier","classification","multimodal","document","support vector machines"],"dc:title":["An Exploration of Multimodal Document Classification Strategies"],"thesis:degree_discipline":["Electrical & Computer Engr"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["M.S."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:25:23Z"}