{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/97736"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/97736","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Beyond topic-based representations for text mining","abstract":"\"A massive amount of online information is natural language text: newspapers, blog articles, forum posts and comments, tweets, scientific literature, government documents, and more. While in general, all kinds of online information is useful, textual information is especially important—it is the most natural, most common, and most expressive form of information. Text representation plays a critical role in application tasks like classification or information retrieval since the quality of the underlying feature space directly impacts each task's performance. Because of this importance, many different approaches have been developed for generating text representations. By far, the most common way to generate features is to segment text into words and record their n-grams. While simple term features perform relatively well in topic-based tasks, not all downstream applications are of a topical nature and can be captured by words alone. For example, determining the native language of an English essay writer will depend on more than just word choice. Competing methods to topic-based representations (such as neural networks) are often not interpretable or rely on massive amounts of training data. This thesis proposes three novel contributions to generate and analyze a large space of non-topical features. First, structural parse tree features are solely based on structural properties of a parse tree by ignoring all of the syntactic categories in the tree. An important advantage of these \"\"skeletons\"\" over regular syntactic features is that they can capture global tree structures without causing problems of data sparseness or overfitting. Second, SyntacticDiff explicitly captures differences in a text document with respect to a reference corpus, creating features that are easily explained as weighted word edit differences. These edit features are especially useful since they are derived from information not present in the current document, capturing a type of comparative feature. Third, Cross-Context Lexical Analysis is a general framework for analyzing similarities and differences in both term meaning and representation with respect to different, potentially overlapping partitions of a text collection. The representations analyzed by CCLA are not limited to topic-based features.\"","abstract_html":""A massive amount of online information is natural language text: newspapers, blog articles, forum posts and comments, tweets, scientific literature, government documents, and more. While in general, all kinds of online information is useful, textual information is especially important—it is the most natural, most common, and most expressive form of information. Text representation plays a critical role in application tasks like classification or information retrieval since the quality of the underlying feature space directly impacts each task's performance. Because of this importance, many different approaches have been developed for generating text representations. By far, the most common way to generate features is to segment text into words and record their n-grams. While simple term features perform relatively well in topic-based tasks, not all downstream applications are of a topical nature and can be captured by words alone. For example, determining the native language of an English essay writer will depend on more than just word choice. Competing methods to topic-based representations (such as neural networks) are often not interpretable or rely on massive amounts of training data. This thesis proposes three novel contributions to generate and analyze a large space of non-topical features. First, structural parse tree features are solely based on structural properties of a parse tree by ignoring all of the syntactic categories in the tree. An important advantage of these ""skeletons"" over regular syntactic features is that they can capture global tree structures without causing problems of data sparseness or overfitting. Second, SyntacticDiff explicitly captures differences in a text document with respect to a reference corpus, creating features that are easily explained as weighted word edit differences. These edit features are especially useful since they are derived from information not present in the current document, capturing a type of comparative feature. Third, Cross-Context Lexical Analysis is a general framework for analyzing similarities and differences in both term meaning and representation with respect to different, potentially overlapping partitions of a text collection. The representations analyzed by CCLA are not limited to topic-based features."","abstract_has_math":false,"creators":["Massung, Sean Alexander"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Zhai, ChengXiang","Hockenmaier, Julia","Roth, Dan","LaValle, Steve","Mei, Qiaozhu"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2017,"date_issued":"2017-08-10T20:33:07Z","date_published":"2017-08-10T20:33:07Z","updated_at":"2026-07-22T22:24:34Z","subjects":["Text mining","Text representation","Feature representation","Natural language processing"],"languages":["en"],"rights":["Copyright 2017 Sean Massung"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/97736","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Zhai, ChengXiang","Hockenmaier, Julia","Roth, Dan","LaValle, Steve","Mei, Qiaozhu"]},{"key":"dc:creator","label":"Author","values":["Massung, Sean Alexander"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2017-08-10T20:33:07Z","2019-08-11T09:15:28Z","2017-04-19","2017-05"]},{"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":["Text mining","Text representation","Feature representation","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 2017 Sean Massung"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/97736"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["\"A massive amount of online information is natural language text: newspapers, blog articles, forum posts and comments, tweets, scientific literature, government documents, and more. While in general, all kinds of online information is useful, textual information is especially important—it is the most natural, most common, and most expressive form of information. Text representation plays a critical role in application tasks like classification or information retrieval since the quality of the underlying feature space directly impacts each task's performance. Because of this importance, many different approaches have been developed for generating text representations. By far, the most common way to generate features is to segment text into words and record their n-grams. While simple term features perform relatively well in topic-based tasks, not all downstream applications are of a topical nature and can be captured by words alone. For example, determining the native language of an English essay writer will depend on more than just word choice. Competing methods to topic-based representations (such as neural networks) are often not interpretable or rely on massive amounts of training data. This thesis proposes three novel contributions to generate and analyze a large space of non-topical features. First, structural parse tree features are solely based on structural properties of a parse tree by ignoring all of the syntactic categories in the tree. An important advantage of these \"\"skeletons\"\" over regular syntactic features is that they can capture global tree structures without causing problems of data sparseness or overfitting. Second, SyntacticDiff explicitly captures differences in a text document with respect to a reference corpus, creating features that are easily explained as weighted word edit differences. These edit features are especially useful since they are derived from information not present in the current document, capturing a type of comparative feature. Third, Cross-Context Lexical Analysis is a general framework for analyzing similarities and differences in both term meaning and representation with respect to different, potentially overlapping partitions of a text collection. The representations analyzed by CCLA are not limited to topic-based features.\"","Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2019-05-01","The student, Sean Massung, accepted the attached license on 2017-04-18 at 22:05.","The student, Sean Massung, submitted this Dissertation for approval on 2017-04-18 at 22:15.","This Dissertation was approved for publication on 2017-04-19 at 12:14.","DSpace SAF Submission Ingestion Package generated from Vireo submission #10878 on 2017-08-10 at 15:06:12","Made available in DSpace on 2017-08-10T20:33:07Z (GMT). No. of bitstreams: 2 MASSUNG-DISSERTATION-2017.pdf: 529424 bytes, checksum: ebff34c21752c005367c99688ed6a1b7 (MD5) LICENSE.txt: 4209 bytes, checksum: 5296c201716594c1c3e3419e99e8a515 (MD5) Previous issue date: 2017-04-19","Embargo set by: Colleen Fallaw for item 102789 Lift date: 2019-08-10T21:27:21Z Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD system","U of I Only Restriction Lifted for Item 102789 on 2019-08-11T09:15:28Z."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Beyond topic-based representations for text mining"]}]}],"canonical_facts":{"dc:contributor":["Zhai, ChengXiang","Hockenmaier, Julia","Roth, Dan","LaValle, Steve","Mei, Qiaozhu"],"dc:creator":["Massung, Sean Alexander"],"dc:date":["2017-08-10T20:33:07Z","2019-08-11T09:15:28Z","2017-04-19","2017-05"],"dc:description":["\"A massive amount of online information is natural language text: newspapers, blog articles, forum posts and comments, tweets, scientific literature, government documents, and more. While in general, all kinds of online information is useful, textual information is especially important—it is the most natural, most common, and most expressive form of information. Text representation plays a critical role in application tasks like classification or information retrieval since the quality of the underlying feature space directly impacts each task's performance. Because of this importance, many different approaches have been developed for generating text representations. By far, the most common way to generate features is to segment text into words and record their n-grams. While simple term features perform relatively well in topic-based tasks, not all downstream applications are of a topical nature and can be captured by words alone. For example, determining the native language of an English essay writer will depend on more than just word choice. Competing methods to topic-based representations (such as neural networks) are often not interpretable or rely on massive amounts of training data. This thesis proposes three novel contributions to generate and analyze a large space of non-topical features. First, structural parse tree features are solely based on structural properties of a parse tree by ignoring all of the syntactic categories in the tree. An important advantage of these \"\"skeletons\"\" over regular syntactic features is that they can capture global tree structures without causing problems of data sparseness or overfitting. Second, SyntacticDiff explicitly captures differences in a text document with respect to a reference corpus, creating features that are easily explained as weighted word edit differences. These edit features are especially useful since they are derived from information not present in the current document, capturing a type of comparative feature. Third, Cross-Context Lexical Analysis is a general framework for analyzing similarities and differences in both term meaning and representation with respect to different, potentially overlapping partitions of a text collection. The representations analyzed by CCLA are not limited to topic-based features.\"","Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2019-05-01","The student, Sean Massung, accepted the attached license on 2017-04-18 at 22:05.","The student, Sean Massung, submitted this Dissertation for approval on 2017-04-18 at 22:15.","This Dissertation was approved for publication on 2017-04-19 at 12:14.","DSpace SAF Submission Ingestion Package generated from Vireo submission #10878 on 2017-08-10 at 15:06:12","Made available in DSpace on 2017-08-10T20:33:07Z (GMT). No. of bitstreams: 2 MASSUNG-DISSERTATION-2017.pdf: 529424 bytes, checksum: ebff34c21752c005367c99688ed6a1b7 (MD5) LICENSE.txt: 4209 bytes, checksum: 5296c201716594c1c3e3419e99e8a515 (MD5) Previous issue date: 2017-04-19","Embargo set by: Colleen Fallaw for item 102789 Lift date: 2019-08-10T21:27:21Z Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD system","U of I Only Restriction Lifted for Item 102789 on 2019-08-11T09:15:28Z."],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/97736"],"dc:language":["en"],"dc:rights":["Copyright 2017 Sean Massung"],"dc:subject":["Text mining","Text representation","Feature representation","Natural language processing"],"dc:title":["Beyond topic-based representations for text mining"],"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":"2026-07-22T22:24:34Z"}