{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,17]],"date-time":"2026-05-17T09:08:13Z","timestamp":1779008893303,"version":"3.51.4"},"reference-count":145,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,7,23]],"date-time":"2024-07-23T00:00:00Z","timestamp":1721692800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100020884","name":"Agencia Nacional de Investigaci\u00f3n y Desarrollo","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100020884","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>Deep learning models have achieved state-of-the-art performance for text classification in the last two decades. However, this has come at the expense of models becoming less understandable, limiting their application scope in high-stakes domains. The increased interest in explainability has resulted in many proposed forms of explanation. Nevertheless, recent studies have shown that<jats:italic>rationales<\/jats:italic>, or language explanations, are more intuitive and human-understandable, especially for non-technical stakeholders. This survey provides an overview of the progress the community has achieved thus far in rationalization approaches for text classification. We first describe and compare techniques for producing extractive and abstractive rationales. Next, we present various rationale-annotated data sets that facilitate the training and evaluation of rationalization models. Then, we detail proxy-based and human-grounded metrics to evaluate machine-generated rationales. Finally, we outline current challenges and encourage directions for future work.<\/jats:p>","DOI":"10.3389\/frai.2024.1363531","type":"journal-article","created":{"date-parts":[[2024,7,23]],"date-time":"2024-07-23T05:10:23Z","timestamp":1721711423000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":5,"title":["From outputs to insights: a survey of rationalization approaches for explainable text classification"],"prefix":"10.3389","volume":"7","author":[{"given":"Erick","family":"Mendez Guzman","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Viktor","family":"Schlegel","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Riza","family":"Batista-Navarro","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2024,7,23]]},"reference":[{"key":"B1","doi-asserted-by":"crossref","first-page":"163","DOI":"10.1007\/978-1-4614-3223-4_6","article-title":"\u201cA survey of text classification algorithms,\u201d","volume-title":"Mining Text Data","author":"Aggarwal","year":"2012"},{"key":"B2","doi-asserted-by":"publisher","first-page":"105","DOI":"10.1609\/aimag.v35i4.2513","article-title":"Power to the people: the role of humans in interactive machine learning","volume":"35","author":"Amershi","year":"2014","journal-title":"Ai Mag"},{"key":"B3","doi-asserted-by":"publisher","first-page":"176","DOI":"10.1162\/tacl_a_00540","article-title":"Feelingblue: a corpus for understanding the emotional connotation of color in context","volume":"11","author":"Ananthram","year":"2023","journal-title":"Trans. Assoc. Comput. Linguist"},{"key":"B4","first-page":"5868","article-title":"\u201cMarta: leveraging human rationales for explainable text classification,\u201d","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 35","author":"Arous","year":"2021"},{"key":"B5","doi-asserted-by":"publisher","first-page":"82","DOI":"10.1016\/j.inffus.2019.12.012","article-title":"Explainable artificial intelligence (XAI): concepts, taxonomies, opportunities and challenges toward responsible AI","volume":"58","author":"Arrieta","year":"2020","journal-title":"Inform. Fus"},{"key":"B6","doi-asserted-by":"crossref","first-page":"283","DOI":"10.18653\/v1\/2023.acl-short.25","article-title":"\u201cFaithfulness tests for natural language explanations,\u201d","volume-title":"Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)","author":"Atanasova","year":"2023"},{"key":"B7","doi-asserted-by":"crossref","first-page":"7352","DOI":"10.18653\/v1\/2020.acl-main.656","article-title":"\u201cGenerating fact checking explanations,\u201d","volume-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics","author":"Atanasova","year":"2020"},{"key":"B8","doi-asserted-by":"crossref","first-page":"155","DOI":"10.1007\/978-3-031-51518-7","article-title":"\u201cA diagnostic study of explainability techniques for text classification,\u201d","volume-title":"Accountable and Explainable Methods for Complex Reasoning over Text","author":"Atanasova","year":"2024"},{"key":"B9","article-title":"\u201cNeural machine translation by jointly learning to align and translate,\u201d","author":"Bahdanau","year":"2015","journal-title":"3rd International Conference on Learning Representations, ICLR 2015"},{"key":"B10","doi-asserted-by":"crossref","first-page":"1903","DOI":"10.18653\/v1\/D18-1216","article-title":"\u201cDeriving machine attention from human rationales,\u201d","volume-title":"Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing","author":"Bao","year":"2018"},{"key":"B11","doi-asserted-by":"crossref","first-page":"3214","DOI":"10.18653\/v1\/2020.coling-main.286","article-title":"\u201cRANCC: rationalizing neural networks via concept clustering,\u201d","volume-title":"Proceedings of the 28th International Conference on Computational Linguistics","author":"Bashier","year":"2020"},{"key":"B12","doi-asserted-by":"crossref","first-page":"2963","DOI":"10.18653\/v1\/P19-1284","article-title":"\u201cInterpretable neural predictions with differentiable binary variables,\u201d","volume-title":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics","author":"Bastings","year":"2019"},{"key":"B13","doi-asserted-by":"crossref","first-page":"149","DOI":"10.18653\/v1\/2020.blackboxnlp-1.14","article-title":"\u201cThe elephant in the interpretability room: why use attention as explanation when we have saliency methods?,\u201d","volume-title":"Proceedings of the Third BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP","author":"Bastings","year":"2020"},{"key":"B14","doi-asserted-by":"publisher","first-page":"688969","DOI":"10.3389\/fdata.2021.688969","article-title":"Principles and practice of explainable machine learning","volume":"4","author":"Belle","year":"2021","journal-title":"Front. Big Data"},{"key":"B15","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2109.08259","article-title":"Self-training with few-shot rationalization: teacher explanations aid student in few-shot NLU","author":"Bhat","year":"2021","journal-title":"arXiv preprint arXiv:2109.08259"},{"key":"B16","doi-asserted-by":"publisher","first-page":"149","DOI":"10.1007\/s10506-020-09270-4","article-title":"Legal requirements on explainability in machine learning","volume":"29","author":"Bibal","year":"2021","journal-title":"Artif. Intell. Law"},{"key":"B17","first-page":"440","article-title":"\u201cBiographies, bollywood, boom-boxes and blenders: domain adaptation for sentiment classification,\u201d","volume-title":"Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics","author":"Blitzer","year":"2007"},{"key":"B18","doi-asserted-by":"publisher","first-page":"1877","DOI":"10.48550\/arXiv.2005.14165","article-title":"Language models are few-shot learners","volume":"33","author":"Brown","year":"2020","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B19","doi-asserted-by":"publisher","first-page":"245","DOI":"10.48550\/arXiv.2011.07876","article-title":"A survey on the explainability of supervised machine learning","volume":"70","author":"Burkart","year":"2021","journal-title":"J. Artif. Intell. Res"},{"key":"B20","doi-asserted-by":"publisher","first-page":"1193","DOI":"10.48550\/arXiv.1812.01193","article-title":"e-SNLI: natural language inference with natural language explanations","volume":"31","author":"Camburu","year":"2018","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B21","doi-asserted-by":"crossref","first-page":"4157","DOI":"10.18653\/v1\/2020.acl-main.382","article-title":"\u201cMake up your mind! adversarial generation of inconsistent natural language explanations,\u201d","volume-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics","author":"Camburu","year":"2020"},{"key":"B22","doi-asserted-by":"crossref","first-page":"3497","DOI":"10.18653\/v1\/D18-1386","article-title":"\u201cExtractive adversarial networks: high-recall explanations for identifying personal attacks in social media posts,\u201d","volume-title":"Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing","author":"Carton","year":"2018"},{"key":"B23","doi-asserted-by":"crossref","first-page":"9294","DOI":"10.18653\/v1\/2020.emnlp-main.747","article-title":"\u201cEvaluating and characterizing human rationales,\u201d","volume-title":"Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)","author":"Carton","year":"2020"},{"key":"B24","first-page":"2867","article-title":"\u201cUNIREX: a unified learning framework for language model rationale extraction,\u201d","volume-title":"International Conference on Machine Learning","author":"Chan","year":"2022"},{"key":"B25","doi-asserted-by":"publisher","first-page":"12853","DOI":"10.48550\/arXiv.1910.12853","article-title":"A game theoretic approach to class-wise selective rationalization","volume":"32","author":"Chang","year":"2019","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B26","first-page":"3792","article-title":"\u201cCan rationalization improve robustness?,\u201d","volume-title":"2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL 2022","author":"Chen","year":"2022"},{"key":"B27","doi-asserted-by":"crossref","first-page":"5578","DOI":"10.18653\/v1\/2020.acl-main.494","article-title":"\u201cGenerating hierarchical explanations on text classification via feature interaction detection,\u201d","author":"Chen","year":"2020","journal-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics"},{"key":"B28","first-page":"15586","article-title":"\u201cREX: reasoning-aware and grounded explanation,\u201d","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Chen","year":"2022"},{"key":"B29","doi-asserted-by":"crossref","first-page":"4682","DOI":"10.18653\/v1\/2023.findings-emnlp.310","article-title":"\u201cZARA: improving few-shot self-rationalization for small language models,\u201d","volume-title":"Findings of the Association for Computational Linguistics: EMNLP 2023","author":"Chen","year":"2023"},{"key":"B30","first-page":"10545","article-title":"\u201cFlexible instance-specific rationalization of NLP models,\u201d","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 36","author":"Chrysostomou","year":"2022"},{"key":"B31","first-page":"447","article-title":"\u201cA survey of the state of explainable AI for natural language processing,\u201d","volume-title":"Proceedings of the 1st Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 10th International Joint Conference on Natural Language Processing","author":"Danilevsky","year":"2020"},{"key":"B32","first-page":"4171","article-title":"\u201cBERT: pre-training of deep bidirectional transformers for language understanding,\u201d","volume-title":"Proceedings of NAACL-HLT","author":"Devlin","year":"2019"},{"key":"B33","doi-asserted-by":"crossref","first-page":"4443","DOI":"10.18653\/v1\/2020.acl-main.408","article-title":"\u201cERASER: a benchmark to evaluate rationalized NLP models,\u201d","volume-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics","author":"DeYoung","year":"2020"},{"key":"B34","doi-asserted-by":"publisher","first-page":"3197","DOI":"10.48550\/arXiv.1905.03197","article-title":"Unified language model pre-training for natural language understanding and generation","volume":"32","author":"Dong","year":"2019","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B35","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1702.08608","article-title":"Towards a rigorous science of interpretable machine learning","author":"Doshi-Velez","year":"2017","journal-title":"arXiv preprint arXiv:1702.08608"},{"key":"B36","doi-asserted-by":"crossref","first-page":"263","DOI":"10.1145\/3301275.3302316","article-title":"\u201cAutomated rationale generation: a technique for explainable AI and its effects on human perceptions,\u201d","volume-title":"Proceedings of the 24th International Conference on Intelligent User Interfaces","author":"Ehsan","year":"2019"},{"key":"B37","doi-asserted-by":"crossref","first-page":"80","DOI":"10.1109\/DSAA.2018.00018","article-title":"\u201cExplaining explanations: an overview of interpretability of machine learning,\u201d","volume-title":"2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA)","author":"Gilpin","year":"2018"},{"key":"B38","doi-asserted-by":"crossref","first-page":"6534","DOI":"10.18653\/v1\/2021.emnlp-main.525","article-title":"\u201cSECTRA: sparse structured text rationalization,\u201d","volume-title":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","author":"Guerreiro","year":"2021"},{"key":"B39","doi-asserted-by":"publisher","first-page":"37","DOI":"10.1126\/scirobotics.aay7120","article-title":"Xai\u2013explainable artificial intelligence","volume":"4","author":"Gunning","year":"2019","journal-title":"Sci. Robot"},{"key":"B40","doi-asserted-by":"crossref","first-page":"19","DOI":"10.1109\/STC55697.2022.00012","article-title":"\u201cEXCLAIM: explainable neural claim verification using rationalization,\u201d","volume-title":"2022 IEEE 29th Annual Software Technology Conference (STC)","author":"Gurrapu","year":"2022"},{"key":"B41","doi-asserted-by":"publisher","first-page":"1225093","DOI":"10.3389\/frai.2023.1225093","article-title":"Rationalization for explainable NLP: a survey","volume":"6","author":"Gurrapu","year":"2023","journal-title":"Front. Artif. Intell"},{"key":"B42","first-page":"19","article-title":"\u201cResults of the active learning challenge,\u201d","volume-title":"Active Learning and Experimental Design Workshop in Conjunction With AISTATS 2010. JMLR Workshop and Conference Proceedings","author":"Guyon","year":"2011"},{"key":"B43","doi-asserted-by":"crossref","first-page":"493","DOI":"10.18653\/v1\/K19-1046","article-title":"\u201cA richly annotated corpus for different tasks in automated fact-checking,\u201d","volume-title":"Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL)","author":"Hanselowski","year":"2019"},{"key":"B44","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/2022.lnls-1.5","article-title":"\u201cA survey on improving NLP models with human explanations,\u201d","volume-title":"ACL Workshop on Learning with Natural Language Supervision","author":"Hartmann","year":"2022"},{"key":"B45","doi-asserted-by":"crossref","first-page":"5540","DOI":"10.18653\/v1\/2020.acl-main.491","article-title":"\u201cEvaluating Explainable AI: Which Algorithmic Explanations Help Users Predict Model Behavior?\u201d","volume-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics","author":"Hase","year":"2020"},{"key":"B46","doi-asserted-by":"crossref","first-page":"29","DOI":"10.18653\/v1\/2022.lnls-1.4","article-title":"\u201cWhen can models learn from explanations? A formal framework for understanding the roles of explanation data,\u201d","volume-title":"Proceedings of the First Workshop on Learning with Natural Language Supervision","author":"Hase","year":"2022"},{"key":"B47","doi-asserted-by":"crossref","first-page":"4351","DOI":"10.18653\/v1\/2020.findings-emnlp.390","article-title":"\u201cLeakage-adjusted simulatability: can models generate non-trivial explanations of their behavior in natural language?\u201d","volume-title":"Findings of the Association for Computational Linguistics: EMNLP 2020","author":"Hase","year":"2020"},{"key":"B48","doi-asserted-by":"crossref","first-page":"6323","DOI":"10.18653\/v1\/2021.emnlp-main.510","article-title":"\u201cDoes bert learn as humans perceive? Understanding linguistic styles through lexica,\u201d","volume-title":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","author":"Hayati","year":"2021"},{"key":"B49","article-title":"\u201cDEBERTA: decoding-enhanced bert with disentangled attention,\u201d","volume-title":"International Conference on Learning Representations","author":"He","year":"2020"},{"key":"B50","doi-asserted-by":"crossref","first-page":"1735","DOI":"10.1162\/neco.1997.9.8.1735","article-title":"Long short-term memory","volume":"9","author":"Hochreiter","year":"1997","journal-title":"Neural Comput"},{"key":"B51","doi-asserted-by":"publisher","first-page":"294","DOI":"10.48550\/arXiv.2006.01067","article-title":"Aligning faithful interpretations with their social attribution","volume":"9","author":"Jacovi","year":"2021","journal-title":"Trans. Assoc. Comput. Linguist"},{"key":"B52","first-page":"3543","article-title":"\u201cAttention is not explanation,\u201d","volume-title":"Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)","author":"Jain","year":"2019"},{"key":"B53","doi-asserted-by":"crossref","first-page":"4459","DOI":"10.18653\/v1\/2020.acl-main.409","article-title":"\u201cLearning to faithfully rationalize by construction,\u201d","volume-title":"58th Annual Meeting of the Association for Computational Linguistics, ACL 2020","author":"Jain","year":"2020"},{"key":"B54","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/2023.acl-long.59","article-title":"\u201cBeing right for whose right reasons?\u201d","volume-title":"The 61st Annual Meeting Of The Association For Computational Linguistics","author":"Jakobsen","year":"2023"},{"key":"B55","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2110.02056","article-title":"Are training resources insufficient? Predict first then explain!","author":"Jang","year":"2021","journal-title":"arXiv preprint arXiv:2110.02056"},{"key":"B56","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3571730","article-title":"Survey of hallucination in natural language generation","volume":"55","author":"Ji","year":"2023","journal-title":"ACM Comput. Surv"},{"key":"B57","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2311.02344","article-title":"You only forward once: prediction and rationalization in a single forward pass","author":"Jiang","year":"2023","journal-title":"arXiv preprint arXiv:2311.02344"},{"key":"B58","first-page":"146","article-title":"\u201cInterpretable rationale augmented charge prediction system,\u201d","volume-title":"Proceedings of the 27th International Conference on Computational Linguistics: System Demonstrations","author":"Jiang","year":"2018"},{"key":"B59","doi-asserted-by":"crossref","first-page":"7103","DOI":"10.18653\/v1\/2023.acl-long.392","article-title":"\u201cAre machine rationales (not) useful to humans? Measuring and improving human utility of free-text rationales,\u201d","volume-title":"Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Joshi","year":"2023"},{"key":"B60","article-title":"\u201cRationale-based human-in-the-loop via supervised attention,\u201d","volume-title":"DaSH@ KDD","author":"Kanchinadam","year":"2020"},{"key":"B61","author":"Kandul","year":"2023","journal-title":"Explainable AI: A Review of the Empirical Literature"},{"key":"B62","doi-asserted-by":"publisher","first-page":"150","DOI":"10.48550\/arXiv.1904.08067","article-title":"Text classification algorithms: a survey","volume":"10","author":"Kowsari","year":"2019","journal-title":"Information"},{"key":"B63","doi-asserted-by":"crossref","first-page":"8730","DOI":"10.18653\/v1\/2020.acl-main.771","article-title":"\u201cNILE : natural language inference with faithful natural language explanations,\u201d","author":"Kumar","year":"2020","journal-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics"},{"key":"B64","doi-asserted-by":"crossref","first-page":"2187","DOI":"10.18653\/v1\/2020.findings-emnlp.198","article-title":"\u201cZero-shot rationalization by multi-task transfer learning from question answering,\u201d","volume-title":"Findings of the Association for Computational Linguistics: EMNLP 2020","author":"Kung","year":"2020"},{"key":"B65","doi-asserted-by":"crossref","first-page":"164","DOI":"10.18653\/v1\/2022.blackboxnlp-1.14","article-title":"\u201cHuman ratings do not reflect downstream utility: a study of free-text explanations for model predictions,\u201d","volume-title":"Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP","author":"Kunz","year":"2022"},{"key":"B66","doi-asserted-by":"crossref","DOI":"10.1609\/aaai.v29i1.9513","article-title":"\u201cRecurrent convolutional neural networks for text classification,\u201d","volume-title":"Twenty-Ninth AAAI Conference on Artificial Intelligence","author":"Lai","year":"2015"},{"key":"B67","doi-asserted-by":"crossref","first-page":"3712","DOI":"10.18653\/v1\/2021.emnlp-main.301","article-title":"\u201cFID-EX: improving sequence-to-sequence models for extractive rationale generation,\u201d","volume-title":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","author":"Lakhotia","year":"2021"},{"key":"B68","article-title":"\u201cALBERT: a lite bert for self-supervised learning of language representations,\u201d","volume-title":"International Conference on Learning Representations","author":"Lan","year":"2019"},{"key":"B69","author":"Lei","year":"2017","journal-title":"Interpretable Neural Models for Natural Language Processing"},{"key":"B70","doi-asserted-by":"crossref","first-page":"107","DOI":"10.18653\/v1\/D16-1011","article-title":"\u201cRationalizing neural predictions,\u201d","volume-title":"Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing","author":"Lei","year":"2016"},{"key":"B71","doi-asserted-by":"crossref","first-page":"5195","DOI":"10.18653\/v1\/D19-1523","article-title":"\u201cHuman-grounded evaluations of explanation methods for text classification,\u201d","volume-title":"Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)","author":"Lertvittayakumjorn","year":"2019"},{"key":"B72","doi-asserted-by":"crossref","first-page":"7871","DOI":"10.18653\/v1\/2020.acl-main.703","article-title":"\u201cBART: denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension,\u201d","author":"Lewis","year":"2020","journal-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics"},{"key":"B73","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2103.14919","article-title":"You can do better! If you elaborate the reason when making prediction","author":"Li","year":"2021","journal-title":"arXiv preprint arXiv:2103.14919"},{"key":"B74","doi-asserted-by":"publisher","first-page":"1283","DOI":"10.48550\/arXiv.1810.00069","article-title":"Adversarial attack and defense: a survey","volume":"11","author":"Liang","year":"2022","journal-title":"Electronics"},{"key":"B75","first-page":"5570","article-title":"\u201cTowards explainable NLP: a generative explanation framework for text classification,\u201d","author":"Liu","year":"","journal-title":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics"},{"key":"B76","doi-asserted-by":"crossref","first-page":"12771","DOI":"10.18653\/v1\/2023.acl-long.715","article-title":"\u201cMGR: multi-generator based rationalization,\u201d","volume-title":"Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Liu","year":"2023"},{"key":"B77","doi-asserted-by":"publisher","first-page":"6954","DOI":"10.48550\/arXiv.2209.08285","article-title":"FR: folded rationalization with a unified encoder","volume":"35","author":"Liu","year":"2022","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B78","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1907.11692","article-title":"RoBERTa: a robustly optimized bert pretraining approach","author":"Liu","year":"","journal-title":"arXiv preprint arXiv:1907.11692"},{"key":"B79","doi-asserted-by":"publisher","first-page":"7874","DOI":"10.48550\/arXiv.1705.07874","article-title":"A unified approach to interpreting model predictions","volume":"30","author":"Lundberg","year":"2017","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B80","doi-asserted-by":"publisher","first-page":"1","DOI":"10.48550\/arXiv.2209.11326","article-title":"Towards faithful model explanation in NLP: a survey","volume":"2024","author":"Lyu","year":"2024","journal-title":"Comput. Linguist"},{"key":"B81","first-page":"322","article-title":"\u201cZero-shot event extraction via transfer learning: challenges and insights,\u201d","volume-title":"Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers)","author":"Lyu","year":"2021"},{"key":"B82","first-page":"142","article-title":"\u201cLearning word vectors for sentiment analysis,\u201d","volume-title":"Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies","author":"Maas","year":"2011"},{"key":"B83","doi-asserted-by":"crossref","first-page":"587","DOI":"10.18653\/v1\/2023.conll-1.40","article-title":"\u201cREFER: an end-to-end rationale extraction framework for explanation regularization,\u201d","volume-title":"Proceedings of the 27th Conference on Computational Natural Language Learning (CoNLL)","author":"Madani","year":"2023"},{"key":"B84","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3546577","article-title":"Post-hoc interpretability for neural NLP: a survey","volume":"55","author":"Madsen","year":"2022","journal-title":"ACM Comput. Surv"},{"key":"B85","doi-asserted-by":"crossref","first-page":"2044","DOI":"10.1109\/BigData55660.2022.10020626","article-title":"\u201cExplainable text classification techniques in legal document review: locating rationales without using human annotated training text snippets,\u201d","volume-title":"2022 IEEE International Conference on Big Data (Big Data)","author":"Mahoney","year":"2022"},{"key":"B86","first-page":"14786","article-title":"\u201cKnowledge-grounded self-rationalization via extractive and natural language explanations,\u201d","volume-title":"International Conference on Machine Learning","author":"Majumder","year":"2022"},{"key":"B87","doi-asserted-by":"crossref","first-page":"410","DOI":"10.18653\/v1\/2022.findings-naacl.31","article-title":"\u201cFew-shot self-rationalization with natural language prompts,\u201d","volume-title":"Findings of the Association for Computational Linguistics: NAACL 2022","author":"Marasovi\u0107","year":"2022"},{"key":"B88","doi-asserted-by":"crossref","first-page":"2810","DOI":"10.18653\/v1\/2020.findings-emnlp.253","article-title":"\u201cNatural Language Rationales With Full-Stack Visual Reasoning: From Pixels to Semantic Frames to Commonsense Graphs,\u201d","volume-title":"Findings of the Association for Computational Linguistics: EMNLP 2020","author":"Marasovi\u0107","year":"2020"},{"key":"B89","doi-asserted-by":"publisher","first-page":"14867","DOI":"10.48550\/arXiv.2012.10289","article-title":"HateXplain: a benchmark dataset for explainable hate speech detection","volume":"35","author":"Mathew","year":"2021","journal-title":"arXiv"},{"key":"B90","doi-asserted-by":"crossref","first-page":"1020","DOI":"10.1109\/ICDM.2012.110","article-title":"\u201cLearning attitudes and attributes from multi-aspect reviews,\u201d","volume-title":"2012 IEEE 12th International Conference on Data Mining","author":"McAuley","year":"2012"},{"key":"B91","first-page":"3610","article-title":"\u201cRaFoLa: a rationale-annotated corpus for detecting indicators of forced labour,\u201d","volume-title":"Proceedings of the Thirteenth Language Resources and Evaluation Conference","author":"Mendez","year":"2022"},{"key":"B92","doi-asserted-by":"publisher","first-page":"462","DOI":"10.48550\/arXiv.2202.04538","article-title":"Generating training data with language models: towards zero-shot language understanding","volume":"35","author":"Meng","year":"2022","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B93","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.artint.2018.07.007","article-title":"Explanation in artificial intelligence: insights from the social sciences","volume":"267","author":"Miller","year":"2019","journal-title":"Artif. Intell"},{"key":"B94","first-page":"9200","article-title":"\u201cAdaptive perturbation-based gradient estimation for discrete latent variable models,\u201d","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 37","author":"Minervini","year":"2023"},{"key":"B95","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1801.05075","article-title":"A human-grounded evaluation benchmark for local explanations of machine learning","author":"Mohseni","year":"2018","journal-title":"arXiv preprint arXiv:1801.05075"},{"key":"B96","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1912.06248","article-title":"General information bottleneck objectives and their applications to machine learning","author":"Mukherjee","year":"2019","journal-title":"arXiv preprint arXiv:1912.06248"},{"key":"B97","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2004.14546","article-title":"WT5?! training text-to-text models to explain their predictions","author":"Narang","year":"2020","journal-title":"arXiv preprint arXiv:2004.14546"},{"key":"B98","first-page":"7348","article-title":"LP-SparseMAP: differentiable relaxed optimization for sparse structured prediction","volume-title":"International Conference on Machine Learning","author":"Niculae","year":"2020"},{"key":"B99","doi-asserted-by":"publisher","first-page":"48","DOI":"10.1016\/j.neucom.2021.03.091","article-title":"A review on the attention mechanism of deep learning","volume":"452","author":"Niu","year":"2021","journal-title":"Neurocomputing"},{"key":"B100","doi-asserted-by":"publisher","first-page":"604","DOI":"10.1109\/TNNLS.2020.2979670","article-title":"A survey of the usages of deep learning for natural language processing","volume":"32","author":"Otter","year":"2020","journal-title":"IEEE Trans. Neural Netw. Learn. Syst"},{"key":"B101","first-page":"18","article-title":"Variational em algorithms for non-gaussian latent variable models","volume":"2005","author":"Palmer","year":"2005","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B102","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/2020.emnlp-main.153","article-title":"\u201cAn information bottleneck approach for controlling conciseness in rationale extraction,\u201d","volume-title":"Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)","author":"Paranjape","year":"2020"},{"key":"B103","first-page":"8779","article-title":"\u201cMultimodal explanations: Justifying decisions and pointing to the evidence,\u201d","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"Park","year":"2018"},{"key":"B104","first-page":"9","article-title":"Language models are unsupervised multitask learners","volume":"1","author":"Radford","year":"2019","journal-title":"OpenAI Blog"},{"key":"B105","doi-asserted-by":"publisher","first-page":"1","DOI":"10.48550\/arXiv.1910.10683","article-title":"Exploring the limits of transfer learning with a unified text-to-text transformer","volume":"21","author":"Raffel","year":"2020","journal-title":"J. Mach. Learn. Res"},{"key":"B106","doi-asserted-by":"crossref","first-page":"4932","DOI":"10.18653\/v1\/P19-1487","article-title":"\u201cExplain yourself! leveraging language models for commonsense reasoning,\u201d","volume-title":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics","author":"Rajani","year":"2019"},{"key":"B107","first-page":"1135","article-title":"\u201cWhy should i trust you?\u201d explaining the predictions of any classifier,\u201d","volume-title":"Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","author":"Ribeiro","year":""},{"key":"B108","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1606.05386","article-title":"Model-agnostic interpretability of machine learning","author":"Ribeiro","year":"","journal-title":"arXiv preprint arXiv:1606.05386"},{"key":"B109","doi-asserted-by":"crossref","first-page":"7403","DOI":"10.18653\/v1\/2022.emnlp-main.501","article-title":"\u201cDoes self-rationalization improve robustness to spurious correlations?\u201d","volume-title":"Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing","author":"Ross","year":"2022"},{"key":"B110","doi-asserted-by":"crossref","first-page":"4596","DOI":"10.18653\/v1\/2020.acl-main.419","article-title":"\u201cHuman attention maps for text classification: do humans and neural networks focus on the same words?\u201d","volume-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics","author":"Sen","year":"2020"},{"key":"B111","article-title":"\u201cDeep inside convolutional networks: visualising image classification models and saliency maps,\u201d","volume-title":"Proceedings of the International Conference on Learning Representations (ICLR)","author":"Simonyan","year":"2014"},{"key":"B112","first-page":"1631","article-title":"\u201cRecursive deep models for semantic compositionality over a sentiment treebank,\u201d","volume-title":"Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing","author":"Socher","year":"2013"},{"key":"B113","doi-asserted-by":"crossref","first-page":"56","DOI":"10.18653\/v1\/W19-4807","article-title":"\u201cDo human rationales improve machine explanations?\u201d","volume-title":"Proceedings of the 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP","author":"Strout","year":"2019"},{"key":"B114","first-page":"3319","article-title":"\u201cAxiomatic attribution for deep networks,\u201d","volume-title":"International Conference on Machine Learning","author":"Sundararajan","year":"2017"},{"key":"B115","doi-asserted-by":"publisher","first-page":"3215","DOI":"10.48550\/arXiv.1409.3215","article-title":"Sequence to sequence learning with neural networks","volume":"27","author":"Sutskever","year":"2014","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B116","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/2022.naacl-main.158","article-title":"\u201cOn the diversity and limits of human explanations,\u201d","volume-title":"Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies","author":"Tan","year":"2022"},{"key":"B117","first-page":"809","article-title":"\u201cFever: a large-scale dataset for fact extraction and verification,\u201d","volume-title":"Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers)","author":"Thorne","year":"2018"},{"key":"B118","doi-asserted-by":"publisher","first-page":"4793","DOI":"10.48550\/arXiv.1907.07374","article-title":"A survey on explainable artificial intelligence (XAI): toward medical XAI","volume":"32","author":"Tjoa","year":"2020","journal-title":"IEEE Trans. Neural Netw. Learn. Syst"},{"key":"B119","doi-asserted-by":"crossref","first-page":"361","DOI":"10.18653\/v1\/2023.bea-1.29","article-title":"\u201cEXASAG: explainable framework for automatic short answer grading,\u201d","volume-title":"Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2023)","author":"Tornqvist","year":"2023"},{"key":"B120","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1909.11218","article-title":"Attention interpretability across NLP tasks","author":"Vashishth","year":"2019","journal-title":"arXiv preprint arXiv:1909.11218"},{"key":"B121","doi-asserted-by":"publisher","first-page":"3762","DOI":"10.48550\/arXiv.1706.03762","article-title":"Attention is all you need","volume":"30","author":"Vaswani","year":"2017","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B122","doi-asserted-by":"crossref","first-page":"1109","DOI":"10.1109\/ICACCI.2017.8125990","article-title":"\u201cA comprehensive study of text classification algorithms,\u201d","volume-title":"2017 International Conference on Advances in Computing, Communications and Informatics (ICACCI)","author":"Vijayan","year":"2017"},{"key":"B123","doi-asserted-by":"crossref","first-page":"354","DOI":"10.1109\/3CBIT57391.2022.00080","article-title":"\u201cRecent development on extractive rationale for model interpretability: a survey,\u201d","volume-title":"2022 International Conference on Cloud Computing, Big Data and Internet of Things (3CBIT)","author":"Wang","year":"2022"},{"key":"B124","doi-asserted-by":"crossref","first-page":"783","DOI":"10.1145\/1835804.1835903","article-title":"\u201cLatent aspect rating analysis on review text data: a rating regression approach,\u201d","author":"Wang","year":"2010","journal-title":"Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining"},{"key":"B125","article-title":"\u201cFinetuned language models are zero-shot learners,\u201d","volume-title":"International Conference on Learning Representations","author":"Wei","year":"2021"},{"key":"B126","article-title":"\u201cTeach me to explain: a review of datasets for explainable natural language processing,\u201d","volume-title":"Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 1)","author":"Wiegreffe","year":"2021"},{"key":"B127","doi-asserted-by":"crossref","first-page":"10266","DOI":"10.18653\/v1\/2021.emnlp-main.804","article-title":"\u201cMeasuring association between labels and free-text rationales,\u201d","volume-title":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","author":"Wiegreffe","year":"2021"},{"key":"B128","doi-asserted-by":"crossref","first-page":"11","DOI":"10.18653\/v1\/D19-1002","article-title":"\u201cAttention is not explanation,\u201d","volume-title":"Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)","author":"Wiegreffe","year":"2019"},{"key":"B129","doi-asserted-by":"crossref","first-page":"229","DOI":"10.1007\/BF00992696","article-title":"Simple statistical gradient-following algorithms for connectionist reinforcement learning","volume":"8","author":"Williams","year":"1992","journal-title":"Machine learn"},{"key":"B130","first-page":"483","article-title":"\u201cMT5: a massively multilingual pre-trained text-to-text transformer,\u201d","volume-title":"Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies","author":"Xue","year":"2021"},{"key":"B131","doi-asserted-by":"crossref","first-page":"14698","DOI":"10.18653\/v1\/2023.acl-long.821","article-title":"\u201cAre human explanations always helpful? Towards objective evaluation of human natural language explanations,\u201d","volume-title":"Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Yao","year":"2023"},{"key":"B132","first-page":"4094","article-title":"\u201cRethinking cooperative rationalization: introspective extraction and complement control,\u201d","author":"Yu","year":"2019","journal-title":"Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)"},{"key":"B133","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2402.10828","article-title":"RAG-Driver: generalisable driving explanations with retrieval-augmented in-context learning in multi-modal large language model","author":"Yuan","year":"2024","journal-title":"arXiv preprint arXiv:2402.10828"},{"key":"B134","first-page":"260","article-title":"\u201cUsing \u201cannotator rationales\u201d to improve machine learning for text categorization,\u201d","volume-title":"Human Language Technologies 2007: The Conference of the North American Chapter of the Association for Computational Linguistics; Proceedings of the Main Conference","author":"Zaidan","year":"2007"},{"key":"B135","first-page":"818","article-title":"\u201cVisualizing and understanding convolutional networks,\u201d","volume-title":"European Conference on Computer Vision","author":"Zeiler","year":"2014"},{"key":"B136","first-page":"957","article-title":"\u201cHuman-like explanation for text classification with limited attention supervision,\u201d","volume-title":"2021 IEEE International Conference on Big Data","author":"Zhang","year":""},{"key":"B137","first-page":"10887","article-title":"\u201cSample efficient reinforcement learning with reinforce,\u201d","author":"Zhang","year":"","journal-title":"Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 35"},{"key":"B138","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2403.11150","article-title":"Training a small emotional vision language model for visual art comprehension","author":"Zhang","year":"2024","journal-title":"arXiv preprint arXiv:2403.11150"},{"key":"B139","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3374217","article-title":"Adversarial attacks on deep-learning models in natural language processing: a survey","volume":"11","author":"Zhang","year":"2020","journal-title":"ACM Trans. Intell. Syst. Technol"},{"key":"B140","doi-asserted-by":"crossref","first-page":"795","DOI":"10.18653\/v1\/D16-1076","article-title":"\u201cRationale-augmented convolutional neural networks for text classification,\u201d","volume-title":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing, Vol. 2016","author":"Zhang","year":"2016"},{"key":"B141","doi-asserted-by":"publisher","first-page":"1029","DOI":"10.48550\/arXiv.2309.01029","article-title":"Explainability for large language models: a survey","volume":"2023","author":"Zhao","year":"2023","journal-title":"ACM Trans. Intell. Syst. Technol"},{"key":"B142","doi-asserted-by":"crossref","first-page":"14532","DOI":"10.1609\/aaai.v35i16.17708","article-title":"\u201cLIREX: augmenting language inference with relevant explanations,\u201d","author":"Zhao","year":"2021","journal-title":"Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 35"},{"key":"B143","doi-asserted-by":"publisher","first-page":"6803","DOI":"10.48550\/arXiv.2011.05268","article-title":"Towards interpretable natural language understanding with explanations as latent variables","volume":"33","author":"Zhou","year":"2020","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B144","doi-asserted-by":"crossref","first-page":"6743","DOI":"10.18653\/v1\/2023.acl-long.372","article-title":"\u201cFLAME: few-shot learning from natural language explanations,\u201d","volume-title":"Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Zhou","year":"2023"},{"key":"B145","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3529755","article-title":"On the explainability of natural language processing deep models","volume":"55","author":"Zini","year":"2022","journal-title":"ACM Comput. Surv"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2024.1363531\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,11,24]],"date-time":"2024-11-24T16:21:08Z","timestamp":1732465268000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2024.1363531\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,7,23]]},"references-count":145,"alternative-id":["10.3389\/frai.2024.1363531"],"URL":"https:\/\/doi.org\/10.3389\/frai.2024.1363531","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,7,23]]},"article-number":"1363531"}}