{"id":{"repo_id":"mit","oai_identifier":"oai:dspace.mit.edu:1721.1/144758"},"canonical_url":"https://search.dev.ndltd.org/etd/mit/oai:dspace.mit.edu:1721.1/144758","repository":{"repo_id":"mit","name":"MIT","base_url":"https://dspace.mit.edu/oai/request"},"display":{"title":"Self-Training for Natural Language Processing","abstract":"Data annotation is critical for machine learning based natural language processing models. Although many large-scale corpora and standard benchmarks have been annotated and published, they cannot cover all possible applications. As a result, it is difficult to transfer models trained with public corpora to tasks that require domain-specific knowledge, different inference skills, unseen text styles, and explainability. In this thesis, we explore self-training methods for mitigating the data distribution gaps between training and evaluation domains and tasks. In contrast to traditional self-training methods that study the best practice of training models with real data and pseudo labels, we also explore the possibility of automatically generating synthetic data for better explainability, robustness, and domain adaptation performance. We show the performance improvement achieved by our methods on different natural language understanding and generation tasks, including question answering, question generation, and dialog response selection.","abstract_html":"Data annotation is critical for machine learning based natural language processing models. Although many large-scale corpora and standard benchmarks have been annotated and published, they cannot cover all possible applications. As a result, it is difficult to transfer models trained with public corpora to tasks that require domain-specific knowledge, different inference skills, unseen text styles, and explainability. In this thesis, we explore self-training methods for mitigating the data distribution gaps between training and evaluation domains and tasks. In contrast to traditional self-training methods that study the best practice of training models with real data and pseudo labels, we also explore the possibility of automatically generating synthetic data for better explainability, robustness, and domain adaptation performance. We show the performance improvement achieved by our methods on different natural language understanding and generation tasks, including question answering, question generation, and dialog response selection.","abstract_has_math":false,"creators":["Luo, Hongyin"],"institution":"Massachusetts Institute of Technology","degree_name":"Doctoral","degree_level":null,"degree_discipline":null,"degree_department":"Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science","school":null,"contributors":[],"advisors":["Glass, James R."],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-05","date_published":"2022-05","updated_at":"2026-07-22T22:21:12Z","subjects":[],"languages":[],"rights":["In Copyright - Educational Use Permitted","Copyright MIT"],"rights_urls":["http://rightsstatements.org/page/InC-EDU/1.0/"],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/1721.1/144758","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Glass, James R."]},{"key":"dc:contributor.department","label":"Department","values":["Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science"]},{"key":"dc:creator","label":"Author","values":["Luo, Hongyin"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2022-08-29T16:09:38Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2022-08-29T16:09:38Z"]},{"key":"dc:date.issued","label":"Date","values":["2022-05"]},{"key":"dc:publisher","label":"Institution","values":["Massachusetts Institute of Technology"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctoral","Doctor of Philosophy"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["In Copyright - Educational Use Permitted","Copyright MIT"]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://rightsstatements.org/page/InC-EDU/1.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/1721.1/144758"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Data annotation is critical for machine learning based natural language processing models. Although many large-scale corpora and standard benchmarks have been annotated and published, they cannot cover all possible applications. As a result, it is difficult to transfer models trained with public corpora to tasks that require domain-specific knowledge, different inference skills, unseen text styles, and explainability. In this thesis, we explore self-training methods for mitigating the data distribution gaps between training and evaluation domains and tasks. In contrast to traditional self-training methods that study the best practice of training models with real data and pseudo labels, we also explore the possibility of automatically generating synthetic data for better explainability, robustness, and domain adaptation performance. We show the performance improvement achieved by our methods on different natural language understanding and generation tasks, including question answering, question generation, and dialog response selection."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["Ph.D."]},{"key":"dc:title","label":"Title","values":["Self-Training for Natural Language Processing"]}]}],"canonical_facts":{"dc:contributor.advisor":["Glass, James R."],"dc:contributor.department":["Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science"],"dc:creator":["Luo, Hongyin"],"dc:date.accessioned":["2022-08-29T16:09:38Z"],"dc:date.available":["2022-08-29T16:09:38Z"],"dc:date.issued":["2022-05"],"dc:description.abstract":["Data annotation is critical for machine learning based natural language processing models. Although many large-scale corpora and standard benchmarks have been annotated and published, they cannot cover all possible applications. As a result, it is difficult to transfer models trained with public corpora to tasks that require domain-specific knowledge, different inference skills, unseen text styles, and explainability. In this thesis, we explore self-training methods for mitigating the data distribution gaps between training and evaluation domains and tasks. In contrast to traditional self-training methods that study the best practice of training models with real data and pseudo labels, we also explore the possibility of automatically generating synthetic data for better explainability, robustness, and domain adaptation performance. We show the performance improvement achieved by our methods on different natural language understanding and generation tasks, including question answering, question generation, and dialog response selection."],"dc:description.degree":["Ph.D."],"dc:identifier.uri":["https://hdl.handle.net/1721.1/144758"],"dc:publisher":["Massachusetts Institute of Technology"],"dc:rights":["In Copyright - Educational Use Permitted","Copyright MIT"],"dc:rights.uri":["http://rightsstatements.org/page/InC-EDU/1.0/"],"dc:title":["Self-Training for Natural Language Processing"],"dc:type":["Thesis"],"thesis:degree_name":["Doctoral","Doctor of Philosophy"]},"updated_at":"2026-07-22T22:21:12Z"}