Zero-Shot Classification by Logical Reasoning on Natural Language Explanations

Humans can classify data of an unseen category by reasoning on its language explanations. This ability is owing to the compositional nature of language: we can combine previously seen attributes to describe the new category. For example, we might describe a sage thrasher as "it has a slim strai...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2023-05
Hauptverfasser: Han, Chi, Pei, Hengzhi, Du, Xinya, Ji, Heng
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Han, Chi
Pei, Hengzhi
Du, Xinya
Ji, Heng
description Humans can classify data of an unseen category by reasoning on its language explanations. This ability is owing to the compositional nature of language: we can combine previously seen attributes to describe the new category. For example, we might describe a sage thrasher as "it has a slim straight relatively short bill, yellow eyes and a long tail", so that others can use their knowledge of attributes "slim straight relatively short bill", "yellow eyes" and "long tail" to recognize a sage thrasher. Inspired by this observation, in this work we tackle zero-shot classification task by logically parsing and reasoning on natural language expla-nations. To this end, we propose the framework CLORE (Classification by LOgical Reasoning on Explanations). While previous methods usually regard textual information as implicit features, CLORE parses explanations into logical structures and then explicitly reasons along thess structures on the input to produce a classification score. Experimental results on explanation-based zero-shot classification benchmarks demonstrate that CLORE is superior to baselines, which we further show mainly comes from higher scores on tasks requiring more logical reasoning. We also demonstrate that our framework can be extended to zero-shot classification on visual modality. Alongside classification decisions, CLORE can provide the logical parsing and reasoning process as a clear form of rationale. Through empirical analysis we demonstrate that CLORE is also less affected by linguistic biases than baselines.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2733851315</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2733851315</sourcerecordid><originalsourceid>FETCH-proquest_journals_27338513153</originalsourceid><addsrcrecordid>eNqNi0EKwjAURIMgWLR3CLgutImx3ZeKi9KFunJTvpLGlJDU_AT09hbxAK6GeW9mQRLGeZFVO8ZWJEUc8zxn-5IJwRPSXaV32fnhAq0NIOpB3yFoZ-ntTVun5mboSQI6q62iM-8gRD_DFqyKoCRtXpMB-z3hhiwHMCjTX67J9tBc6mM2efeMEkM_uujtrHpWcl6JgheC_7f6ANMhPkA</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2733851315</pqid></control><display><type>article</type><title>Zero-Shot Classification by Logical Reasoning on Natural Language Explanations</title><source>Freely Accessible Journals</source><creator>Han, Chi ; Pei, Hengzhi ; Du, Xinya ; Ji, Heng</creator><creatorcontrib>Han, Chi ; Pei, Hengzhi ; Du, Xinya ; Ji, Heng</creatorcontrib><description>Humans can classify data of an unseen category by reasoning on its language explanations. This ability is owing to the compositional nature of language: we can combine previously seen attributes to describe the new category. For example, we might describe a sage thrasher as "it has a slim straight relatively short bill, yellow eyes and a long tail", so that others can use their knowledge of attributes "slim straight relatively short bill", "yellow eyes" and "long tail" to recognize a sage thrasher. Inspired by this observation, in this work we tackle zero-shot classification task by logically parsing and reasoning on natural language expla-nations. To this end, we propose the framework CLORE (Classification by LOgical Reasoning on Explanations). While previous methods usually regard textual information as implicit features, CLORE parses explanations into logical structures and then explicitly reasons along thess structures on the input to produce a classification score. Experimental results on explanation-based zero-shot classification benchmarks demonstrate that CLORE is superior to baselines, which we further show mainly comes from higher scores on tasks requiring more logical reasoning. We also demonstrate that our framework can be extended to zero-shot classification on visual modality. Alongside classification decisions, CLORE can provide the logical parsing and reasoning process as a clear form of rationale. Through empirical analysis we demonstrate that CLORE is also less affected by linguistic biases than baselines.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Classification ; Cognition &amp; reasoning ; Empirical analysis ; Feathers ; Natural language ; Natural language processing ; Reasoning</subject><ispartof>arXiv.org, 2023-05</ispartof><rights>2023. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>781,785</link.rule.ids></links><search><creatorcontrib>Han, Chi</creatorcontrib><creatorcontrib>Pei, Hengzhi</creatorcontrib><creatorcontrib>Du, Xinya</creatorcontrib><creatorcontrib>Ji, Heng</creatorcontrib><title>Zero-Shot Classification by Logical Reasoning on Natural Language Explanations</title><title>arXiv.org</title><description>Humans can classify data of an unseen category by reasoning on its language explanations. This ability is owing to the compositional nature of language: we can combine previously seen attributes to describe the new category. For example, we might describe a sage thrasher as "it has a slim straight relatively short bill, yellow eyes and a long tail", so that others can use their knowledge of attributes "slim straight relatively short bill", "yellow eyes" and "long tail" to recognize a sage thrasher. Inspired by this observation, in this work we tackle zero-shot classification task by logically parsing and reasoning on natural language expla-nations. To this end, we propose the framework CLORE (Classification by LOgical Reasoning on Explanations). While previous methods usually regard textual information as implicit features, CLORE parses explanations into logical structures and then explicitly reasons along thess structures on the input to produce a classification score. Experimental results on explanation-based zero-shot classification benchmarks demonstrate that CLORE is superior to baselines, which we further show mainly comes from higher scores on tasks requiring more logical reasoning. We also demonstrate that our framework can be extended to zero-shot classification on visual modality. Alongside classification decisions, CLORE can provide the logical parsing and reasoning process as a clear form of rationale. Through empirical analysis we demonstrate that CLORE is also less affected by linguistic biases than baselines.</description><subject>Classification</subject><subject>Cognition &amp; reasoning</subject><subject>Empirical analysis</subject><subject>Feathers</subject><subject>Natural language</subject><subject>Natural language processing</subject><subject>Reasoning</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNi0EKwjAURIMgWLR3CLgutImx3ZeKi9KFunJTvpLGlJDU_AT09hbxAK6GeW9mQRLGeZFVO8ZWJEUc8zxn-5IJwRPSXaV32fnhAq0NIOpB3yFoZ-ntTVun5mboSQI6q62iM-8gRD_DFqyKoCRtXpMB-z3hhiwHMCjTX67J9tBc6mM2efeMEkM_uujtrHpWcl6JgheC_7f6ANMhPkA</recordid><startdate>20230525</startdate><enddate>20230525</enddate><creator>Han, Chi</creator><creator>Pei, Hengzhi</creator><creator>Du, Xinya</creator><creator>Ji, Heng</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20230525</creationdate><title>Zero-Shot Classification by Logical Reasoning on Natural Language Explanations</title><author>Han, Chi ; Pei, Hengzhi ; Du, Xinya ; Ji, Heng</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_27338513153</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Classification</topic><topic>Cognition &amp; reasoning</topic><topic>Empirical analysis</topic><topic>Feathers</topic><topic>Natural language</topic><topic>Natural language processing</topic><topic>Reasoning</topic><toplevel>online_resources</toplevel><creatorcontrib>Han, Chi</creatorcontrib><creatorcontrib>Pei, Hengzhi</creatorcontrib><creatorcontrib>Du, Xinya</creatorcontrib><creatorcontrib>Ji, Heng</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection (ProQuest)</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Han, Chi</au><au>Pei, Hengzhi</au><au>Du, Xinya</au><au>Ji, Heng</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Zero-Shot Classification by Logical Reasoning on Natural Language Explanations</atitle><jtitle>arXiv.org</jtitle><date>2023-05-25</date><risdate>2023</risdate><eissn>2331-8422</eissn><abstract>Humans can classify data of an unseen category by reasoning on its language explanations. This ability is owing to the compositional nature of language: we can combine previously seen attributes to describe the new category. For example, we might describe a sage thrasher as "it has a slim straight relatively short bill, yellow eyes and a long tail", so that others can use their knowledge of attributes "slim straight relatively short bill", "yellow eyes" and "long tail" to recognize a sage thrasher. Inspired by this observation, in this work we tackle zero-shot classification task by logically parsing and reasoning on natural language expla-nations. To this end, we propose the framework CLORE (Classification by LOgical Reasoning on Explanations). While previous methods usually regard textual information as implicit features, CLORE parses explanations into logical structures and then explicitly reasons along thess structures on the input to produce a classification score. Experimental results on explanation-based zero-shot classification benchmarks demonstrate that CLORE is superior to baselines, which we further show mainly comes from higher scores on tasks requiring more logical reasoning. We also demonstrate that our framework can be extended to zero-shot classification on visual modality. Alongside classification decisions, CLORE can provide the logical parsing and reasoning process as a clear form of rationale. Through empirical analysis we demonstrate that CLORE is also less affected by linguistic biases than baselines.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2023-05
issn 2331-8422
language eng
recordid cdi_proquest_journals_2733851315
source Freely Accessible Journals
subjects Classification
Cognition & reasoning
Empirical analysis
Feathers
Natural language
Natural language processing
Reasoning
title Zero-Shot Classification by Logical Reasoning on Natural Language Explanations
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-11T10%3A08%3A15IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Zero-Shot%20Classification%20by%20Logical%20Reasoning%20on%20Natural%20Language%20Explanations&rft.jtitle=arXiv.org&rft.au=Han,%20Chi&rft.date=2023-05-25&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2733851315%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2733851315&rft_id=info:pmid/&rfr_iscdi=true