{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/102868"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/102868","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Reinforced co-learning for semi-supervised ranking","abstract":"\"Learning to rank is vital to information retrieval and recommendation systems. Directly optimizing the listwise evaluation measure such as normalized discounted cumulative gain (NDCG) is an advanced way to learn a ranking model. However, this is only suited for training data with effective labels. In real applications, we are more often faced with the semi-supervised setting that only a partial set of data has labels. In this paper, we propose a co-learning strategy for the semi-supervised ranking problem. Our model has two modules: the classifier module and the reinforcement ranker module. Given a query, the classifier module is trained to classify whether a document is relevant or not. The reinforcement ranker module is trained to give relevance scores on the basis of treating ranking problems as Markov decision processes (MDP). We name our approach \"\"reinforced co-learning\"\" because the two modules are iteratively optimized and affect each other while training. When training the classifier module, we use the reinforcement module to give every candidate a relevance score and sample lower scored documents as irrelevant samples (negative samples). Likewise, in order to train the reinforcement ranker module, we use the classifier module to predict labels in the sequence in order to calculate the combined rewards. The linkage between the two modules is also reflected in the network structure. We add the feature sharing layer, which enables the classifier to distill its intermediate representations to the learning of reinforcement ranker module. Extensive experiments and ablation studies show that both our co-learning strategy and feature sharing can improve semi-supervised ranking problems.\"","abstract_html":""Learning to rank is vital to information retrieval and recommendation systems. Directly optimizing the listwise evaluation measure such as normalized discounted cumulative gain (NDCG) is an advanced way to learn a ranking model. However, this is only suited for training data with effective labels. In real applications, we are more often faced with the semi-supervised setting that only a partial set of data has labels. In this paper, we propose a co-learning strategy for the semi-supervised ranking problem. Our model has two modules: the classifier module and the reinforcement ranker module. Given a query, the classifier module is trained to classify whether a document is relevant or not. The reinforcement ranker module is trained to give relevance scores on the basis of treating ranking problems as Markov decision processes (MDP). We name our approach ""reinforced co-learning"" because the two modules are iteratively optimized and affect each other while training. When training the classifier module, we use the reinforcement module to give every candidate a relevance score and sample lower scored documents as irrelevant samples (negative samples). Likewise, in order to train the reinforcement ranker module, we use the classifier module to predict labels in the sequence in order to calculate the combined rewards. The linkage between the two modules is also reflected in the network structure. We add the feature sharing layer, which enables the classifier to distill its intermediate representations to the learning of reinforcement ranker module. Extensive experiments and ablation studies show that both our co-learning strategy and feature sharing can improve semi-supervised ranking problems."","abstract_has_math":false,"creators":["He, Shibi"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"M.S.","degree_level":"Thesis","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Peng, Jian"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2019,"date_issued":"2019-02-07T20:44:31Z","date_published":"2019-02-07T20:44:31Z","updated_at":"2026-07-22T22:24:42Z","subjects":["Learning to rank","Semi-supervised Learning","Reinforcement Learning","Machine Learning"],"languages":["en"],"rights":["Copyright 2018 Shibi He"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/102868","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Peng, Jian"]},{"key":"dc:creator","label":"Author","values":["He, Shibi"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2019-02-07T20:44:31Z","2021-02-08T10:15:22Z","2018-12-14","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":["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":["Learning to rank","Semi-supervised Learning","Reinforcement Learning","Machine Learning"]}]},{"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 Shibi He"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/102868"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["\"Learning to rank is vital to information retrieval and recommendation systems. Directly optimizing the listwise evaluation measure such as normalized discounted cumulative gain (NDCG) is an advanced way to learn a ranking model. However, this is only suited for training data with effective labels. In real applications, we are more often faced with the semi-supervised setting that only a partial set of data has labels. In this paper, we propose a co-learning strategy for the semi-supervised ranking problem. Our model has two modules: the classifier module and the reinforcement ranker module. Given a query, the classifier module is trained to classify whether a document is relevant or not. The reinforcement ranker module is trained to give relevance scores on the basis of treating ranking problems as Markov decision processes (MDP). We name our approach \"\"reinforced co-learning\"\" because the two modules are iteratively optimized and affect each other while training. When training the classifier module, we use the reinforcement module to give every candidate a relevance score and sample lower scored documents as irrelevant samples (negative samples). Likewise, in order to train the reinforcement ranker module, we use the classifier module to predict labels in the sequence in order to calculate the combined rewards. The linkage between the two modules is also reflected in the network structure. We add the feature sharing layer, which enables the classifier to distill its intermediate representations to the learning of reinforcement ranker module. Extensive experiments and ablation studies show that both our co-learning strategy and feature sharing can improve semi-supervised ranking problems.\"","Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2020-12-01","The student, Shibi He, accepted the attached license on 2018-12-14 at 09:37.","The student, Shibi He, submitted this Thesis for approval on 2018-12-14 at 09:43.","This Thesis was approved for publication on 2018-12-14 at 16:16.","DSpace SAF Submission Ingestion Package generated from Vireo submission #13329 on 2019-02-07 at 14:23:42","Made available in DSpace on 2019-02-07T20:44:31Z (GMT). No. of bitstreams: 2 HE-THESIS-2018.pdf: 1012184 bytes, checksum: efb908528b5b9e945f81f53ce408a54b (MD5) LICENSE.txt: 4205 bytes, checksum: 41d0b4fbda83d46e02a543e8bd64f8bc (MD5) Previous issue date: 2018-12-14","Embargo set by: Seth Robbins for item 109894 Lift date: 2021-02-07T20:44:35Z Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD system","U of I Only Restriction Lifted for Item 109894 on 2021-02-08T10:15:22Z."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Reinforced co-learning for semi-supervised ranking"]}]}],"canonical_facts":{"dc:contributor":["Peng, Jian"],"dc:creator":["He, Shibi"],"dc:date":["2019-02-07T20:44:31Z","2021-02-08T10:15:22Z","2018-12-14","2018-12"],"dc:description":["\"Learning to rank is vital to information retrieval and recommendation systems. Directly optimizing the listwise evaluation measure such as normalized discounted cumulative gain (NDCG) is an advanced way to learn a ranking model. However, this is only suited for training data with effective labels. In real applications, we are more often faced with the semi-supervised setting that only a partial set of data has labels. In this paper, we propose a co-learning strategy for the semi-supervised ranking problem. Our model has two modules: the classifier module and the reinforcement ranker module. Given a query, the classifier module is trained to classify whether a document is relevant or not. The reinforcement ranker module is trained to give relevance scores on the basis of treating ranking problems as Markov decision processes (MDP). We name our approach \"\"reinforced co-learning\"\" because the two modules are iteratively optimized and affect each other while training. When training the classifier module, we use the reinforcement module to give every candidate a relevance score and sample lower scored documents as irrelevant samples (negative samples). Likewise, in order to train the reinforcement ranker module, we use the classifier module to predict labels in the sequence in order to calculate the combined rewards. The linkage between the two modules is also reflected in the network structure. We add the feature sharing layer, which enables the classifier to distill its intermediate representations to the learning of reinforcement ranker module. Extensive experiments and ablation studies show that both our co-learning strategy and feature sharing can improve semi-supervised ranking problems.\"","Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2020-12-01","The student, Shibi He, accepted the attached license on 2018-12-14 at 09:37.","The student, Shibi He, submitted this Thesis for approval on 2018-12-14 at 09:43.","This Thesis was approved for publication on 2018-12-14 at 16:16.","DSpace SAF Submission Ingestion Package generated from Vireo submission #13329 on 2019-02-07 at 14:23:42","Made available in DSpace on 2019-02-07T20:44:31Z (GMT). No. of bitstreams: 2 HE-THESIS-2018.pdf: 1012184 bytes, checksum: efb908528b5b9e945f81f53ce408a54b (MD5) LICENSE.txt: 4205 bytes, checksum: 41d0b4fbda83d46e02a543e8bd64f8bc (MD5) Previous issue date: 2018-12-14","Embargo set by: Seth Robbins for item 109894 Lift date: 2021-02-07T20:44:35Z Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD system","U of I Only Restriction Lifted for Item 109894 on 2021-02-08T10:15:22Z."],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/102868"],"dc:language":["en"],"dc:rights":["Copyright 2018 Shibi He"],"dc:subject":["Learning to rank","Semi-supervised Learning","Reinforcement Learning","Machine Learning"],"dc:title":["Reinforced co-learning for semi-supervised ranking"],"dc:type":["text"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["M.S."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:24:42Z"}