{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,24]],"date-time":"2026-10-24T17:14:48Z","timestamp":1792862088140,"version":"4.5.0"},"reference-count":45,"publisher":"MIT Press","license":[{"start":{"date-parts":[[2022,3,23]],"date-time":"2022-03-23T00:00:00Z","timestamp":1647993600000},"content-version":"vor","delay-in-days":81,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2022,3,25]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>We present mGENRE, a sequence-to- sequence system for the Multilingual Entity Linking (MEL) problem\u2014the task of resolving language-specific mentions to a multilingual Knowledge Base (KB). For a mention in a given language, mGENRE predicts the name of the target entity left-to-right, token-by-token in an autoregressive fashion. The autoregressive formulation allows us to effectively cross-encode mention string and entity names to capture more interactions than the standard dot product between mention and entity vectors. It also enables fast search within a large KB even for mentions that do not appear in mention tables and with no need for large-scale vector indices. While prior MEL works use a single representation for each entity, we match against entity names of as many languages as possible, which allows exploiting language connections between source input and target name. Moreover, in a zero-shot setting on languages with no training data at all, mGENRE treats the target language as a latent variable that is marginalized at prediction time. This leads to over 50% improvements in average accuracy. We show the efficacy of our approach through extensive evaluation including experiments on three popular MEL benchmarks where we establish new state-of-the-art results. Source code available at https:\/\/github.com\/facebookresearch\/GENRE.<\/jats:p>","DOI":"10.1162\/tacl_a_00460","type":"journal-article","created":{"date-parts":[[2022,3,23]],"date-time":"2022-03-23T15:20:08Z","timestamp":1648048808000},"page":"274-290","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":49,"title":["Multilingual Autoregressive Entity Linking"],"prefix":"10.1162","volume":"10","author":[{"given":"Nicola","family":"De Cao","sequence":"first","affiliation":[{"name":"Facebook AI, UK."},{"name":"University of Amsterdam, NL. nicola.decao@gmail.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ledell","family":"Wu","sequence":"additional","affiliation":[{"name":"Beijing Academy of Artificial Intelligence, China. wuyu.ledell@gmail.com; ledell@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kashyap","family":"Popat","sequence":"additional","affiliation":[{"name":"Facebook AI, UK. kpopat@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mikel","family":"Artetxe","sequence":"additional","affiliation":[{"name":"Facebook AI, UK. artetxe@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Naman","family":"Goyal","sequence":"additional","affiliation":[{"name":"Facebook AI, USA. naman@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mikhail","family":"Plekhanov","sequence":"additional","affiliation":[{"name":"Facebook AI, UK. movb@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Luke","family":"Zettlemoyer","sequence":"additional","affiliation":[{"name":"Facebook AI, USA"},{"name":"University of Washington, USA. lsz@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nicola","family":"Cancedda","sequence":"additional","affiliation":[{"name":"Facebook AI, UK. ncan@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sebastian","family":"Riedel","sequence":"additional","affiliation":[{"name":"Facebook AI, UK"},{"name":"University College London, UK. sriedel@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fabio","family":"Petroni","sequence":"additional","affiliation":[{"name":"Facebook AI, UK. fabiopetroni@fb.com"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"281","published-online":{"date-parts":[[2022,3,25]]},"reference":[{"key":"2022033118512939700_bib1","article-title":"Learning to retrieve reasoning paths over\n                        wikipedia graph for question answering","volume-title":"8th\n                        International Conference on Learning Representations, ICLR 2020, Addis\n                        Ababa, Ethiopia, April 26\u201330, 2020","author":"Asai","year":"2020"},{"key":"2022033118512939700_bib2","unstructured":"Giusepppe\n              Attardi\n            \n          .\n                        2015. Wikiextractor. https:\/\/github.com\/attardi\/wikiextractor"},{"key":"2022033118512939700_bib3","doi-asserted-by":"publisher","first-page":"1247","DOI":"10.1145\/1376616.1376746","article-title":"Freebase: A collaboratively created graph\n                        database for structuring human knowledge","volume-title":"Proceedings of the 2008 ACM SIGMOD International Conference on\n                        Management of Data","author":"Bollacker","year":"2008"},{"key":"2022033118512939700_bib4","article-title":"Learning end-to-end goal- oriented\n                        dialog","volume-title":"5th International Conference on Learning\n                        Representations, ICLR 2017, Toulon, France, April 24\u201326, 2017,\n                        Conference Track Proceedings","author":"Bordes","year":"2017"},{"key":"2022033118512939700_bib5","doi-asserted-by":"publisher","first-page":"7833","DOI":"10.18653\/v1\/2020.emnlp-main.630","article-title":"Entity linking in 100\n                        languages","volume-title":"Proceedings of the 2020 Conference on\n                        Empirical Methods in Natural Language Processing (EMNLP)","author":"Botha","year":"2020"},{"key":"2022033118512939700_bib6","first-page":"9","article-title":"Using encyclopedic knowledge for named\n                        entity disambiguation","volume-title":"11th Conference of the\n                        European Chapter of the Association for Computational Linguistics","author":"Bunescu","year":"2006"},{"key":"2022033118512939700_bib7","doi-asserted-by":"publisher","first-page":"216","DOI":"10.18653\/v1\/K17-1023","article-title":"Robust coreference resolution and entity linking on\n                        dialogues: Character identification on TV show transcripts","volume-title":"Proceedings of the 21st Conference on Computational Natural Language\n                        Learning (CoNLL 2017)","author":"Chen","year":"2017"},{"key":"2022033118512939700_bib8","first-page":"1787","volume-title":"Relational inference for wikification","author":"Cheng","year":"2013"},{"key":"2022033118512939700_bib9","doi-asserted-by":"publisher","first-page":"8440","DOI":"10.18653\/v1\/2020.acl-main.747","article-title":"Unsupervised cross-lingual representation\n                        learning at scale","volume-title":"Proceedings of the 58th Annual\n                        Meeting of the Association for Computational Linguistics","author":"Conneau","year":"2020"},{"key":"2022033118512939700_bib10","first-page":"708","article-title":"Large-scale named entity disambiguation\n                        based on Wikipedia data","volume-title":"Proceedings of the 2007\n                        Joint Conference on Empirical Methods in Natural Language Processing and\n                        Computational Natural Language Learning (EMNLP-CoNLL)","author":"Cucerzan","year":"2007"},{"key":"2022033118512939700_bib11","article-title":"Alana v2: Entertaining and informative\n                        open-domain social dialogue using ontologies and entity\n                        linking","volume-title":"Alexa Prize\n                    Proceedings","author":"Curry","year":"2018"},{"key":"2022033118512939700_bib12","doi-asserted-by":"publisher","first-page":"2306","DOI":"10.18653\/v1\/N19-1240","article-title":"Question answering by reasoning across\n                        documents with graph convolutional networks","volume-title":"Proceedings of the 2019 Conference of the North American Chapter of\n                        the Association for Computational Linguistics: Human Language Technologies,\n                        Volume 1 (Long and Short Papers)","author":"Cao","year":"2019"},{"key":"2022033118512939700_bib13","article-title":"Autoregressive entity\n                        retrieval","volume-title":"International Conference on Learning\n                        Representations","author":"Cao","year":"2021"},{"key":"2022033118512939700_bib14","first-page":"4171","article-title":"BERT: Pre-training of deep bidirectional\n                        transformers for language understanding","volume-title":"Proceedings of the 2019 Conference of the North American Chapter of\n                        the Association for Computational Linguistics: Human Language Technologies,\n                        Volume 1 (Long and Short Papers)","author":"Devlin","year":"2019"},{"key":"2022033118512939700_bib15","first-page":"277","volume-title":"Entity disambiguation for knowledge base\n                        population","author":"Dredze","year":"2010"},{"key":"2022033118512939700_bib16","first-page":"2619","volume-title":"Deep joint entity disambiguation with local\n                        neural attention","author":"Ganea","year":"2017"},{"key":"2022033118512939700_bib17","first-page":"782","volume-title":"Robust disambiguation of named entities in\n                        text","author":"Hoffart","year":"2011"},{"key":"2022033118512939700_bib18","article-title":"A memory efficient baseline for open domain\n                        question answering","author":"Izacard","year":"2020","journal-title":"arXiv preprint arXiv:\n                        2012.15156"},{"key":"2022033118512939700_bib19","article-title":"Overview of TAC-KBP2015 tri-lingual entity\n                        discovery and linking","volume-title":"TAC","author":"Ji","year":"2015"},{"key":"2022033118512939700_bib20","article-title":"Adam: A method for stochastic optimization","volume-title":"3rd International Conference on Learning Representations, ICLR 2015,\n                        San Diego, CA, USA, May 7\u20139, 2015, Conference Track\n                        Proceedings","author":"Kingma","year":"2015"},{"key":"2022033118512939700_bib21","first-page":"652","article-title":"BANNER: An executable survey of advances\n                        in biomedical named entity recognition","volume-title":"Pacific\n                        Symposium on Biocomputing 2008, PSB 2008","author":"Leaman","year":"2008"},{"key":"2022033118512939700_bib22","doi-asserted-by":"publisher","first-page":"7871","DOI":"10.18653\/v1\/2020.acl-main.703","article-title":"BART: Denoising sequence-to-sequence\n                        pre-training for natural language generation, translation, and\n                        comprehension","volume-title":"Proceedings of the 58th Annual\n                        Meeting of the Association for Computational Linguistics","author":"Lewis","year":"2020"},{"key":"2022033118512939700_bib23","doi-asserted-by":"publisher","first-page":"1098","DOI":"10.1162\/tacl_a_00415","article-title":"PAQ: 65 million probably-asked questions\n                        and what you can do with them","volume":"9","author":"Lewis","year":"2021"},{"key":"2022033118512939700_bib24","doi-asserted-by":"publisher","first-page":"726","DOI":"10.1162\/tacl_a_00343","article-title":"Multilingual denoising pre-training for\n                        neural machine translation","volume":"8","author":"Liu","year":"2020"},{"key":"2022033118512939700_bib25","first-page":"255","volume-title":"Cross-language entity linking","author":"McNamee","year":"2011"},{"key":"2022033118512939700_bib26","first-page":"86","article-title":"Neurips 2020 efficientqa competition: Systems, analyses and\n                        lessons learned","volume-title":"Proceedings of the NeurIPS 2020\n                        Competition and Demonstration Track, volume 133 of Proceedings of Machine\n                        Learning Research","author":"Min","year":"2021"},{"key":"2022033118512939700_bib27","doi-asserted-by":"publisher","first-page":"2553","DOI":"10.18653\/v1\/D19-1258","volume-title":"Revealing the importance of semantic retrieval\n                        for machine reading at scale","author":"Nie","year":"2019"},{"key":"2022033118512939700_bib28","doi-asserted-by":"publisher","first-page":"48","DOI":"10.18653\/v1\/N19-4009","article-title":"fairseq: A fast, extensible toolkit for sequence\n                        modeling","volume-title":"Proceedings of the 2019 Conference of\n                        the North American Chapter of the Association for Computational Linguistics\n                        (Demonstrations)","author":"Ott","year":"2019"},{"key":"2022033118512939700_bib29","doi-asserted-by":"publisher","first-page":"2523","DOI":"10.18653\/v1\/2021.naacl-main.200","article-title":"KILT: A benchmark for knowledge intensive\n                        language tasks","volume-title":"Proceedings of the 2021\n                        Conference of the North American Chapter of the Association for\n                        Computational Linguistics: Human Language Technologies","author":"Petroni","year":"2021"},{"key":"2022033118512939700_bib30","doi-asserted-by":"publisher","first-page":"2492","DOI":"10.18653\/v1\/2021.acl-long.194","article-title":"OTTers: One-turn topic transitions for\n                        open-domain dialogue","volume-title":"Proceedings of the 59th\n                        Annual Meeting of the Association for Computational Linguistics and the 11th\n                        International Joint Conference on Natural Language Processing (Volume 1:\n                        Long Papers)","author":"Sevegnani","year":"2021"},{"key":"2022033118512939700_bib31","first-page":"5464","article-title":"Neural cross-lingual entity\n                        linking","volume-title":"Proceedings of the Thirty-Second AAAI\n                        Conference on Artificial Intelligence, (AAAI-18), the 30th innovative\n                        Applications of Artificial Intelligence (IAAI-18), and the 8th AAAI\n                        Symposium on Educational Advances in Artificial Intelligence (EAAI-18), New\n                        Orleans, Louisiana, USA, February 2\u20137, 2018","author":"Sil","year":"2018"},{"issue":"56","key":"2022033118512939700_bib32","first-page":"1929","article-title":"Dropout: A simple way to prevent neural\n                        networks from overfitting","volume":"15","author":"Srivastava","year":"2014"},{"key":"2022033118512939700_bib33","first-page":"1017","article-title":"Generating text with recurrent neural\n                        networks","volume-title":"Proceedings of the 28th International\n                        Conference on Machine Learning, ICML 2011, Bellevue, Washington, USA, June\n                        28 - July 2, 2011","author":"Sutskever","year":"2011"},{"key":"2022033118512939700_bib34","first-page":"3104","article-title":"Sequence to sequence learning with neural\n                        networks","author":"Sutskever","year":"2014"},{"key":"2022033118512939700_bib35","doi-asserted-by":"publisher","first-page":"2818","DOI":"10.1109\/CVPR.2016.308","article-title":"Rethinking the inception architecture for\n                        computer vision","volume-title":"2016 IEEE Conference on Computer\n                        Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV, USA, June\n                        27\u201330, 2016","author":"Szegedy","year":"2016"},{"key":"2022033118512939700_bib36","doi-asserted-by":"publisher","first-page":"589","DOI":"10.18653\/v1\/N16-1072","article-title":"Cross-lingual wikification using multilingual\n                        embeddings","volume-title":"Proceedings of the 2016 Conference of\n                        the North American Chapter of the Association for Computational Linguistics:\n                        Human Language Technologies","author":"Tsai","year":"2016"},{"key":"2022033118512939700_bib37","doi-asserted-by":"publisher","first-page":"2486","DOI":"10.18653\/v1\/D18-1270","volume-title":"Joint multilingual supervision for cross-lingual entity\n                        linking","author":"Upadhyay","year":"2018"},{"issue":"10","key":"2022033118512939700_bib38","doi-asserted-by":"publisher","first-page":"78","DOI":"10.1145\/2629489","article-title":"Wikidata: A free collaborative\n                        knowledgebase","volume":"57","author":"Vrande\u010di\u0107","year":"2014","journal-title":"Communications of the ACM"},{"key":"2022033118512939700_bib39","first-page":"438","article-title":"A network-based end-to-end trainable\n                        task-oriented dialogue system","volume-title":"Proceedings of the\n                        15th Conference of the European Chapter of the Association for Computational\n                        Linguistics: Volume 1, Long Papers","author":"Wen","year":"2017"},{"key":"2022033118512939700_bib40","first-page":"4003","volume-title":"CCNet: Extracting high quality monolingual datasets\n                        from web crawl data","author":"Wenzek","year":"2020"},{"key":"2022033118512939700_bib41","doi-asserted-by":"publisher","first-page":"665","DOI":"10.18653\/v1\/P17-1062","volume-title":"Hybrid code networks: Practical and efficient\n                        end-to-end dialog control with supervised and reinforcement\n                        learning","author":"Williams","year":"2017"},{"key":"2022033118512939700_bib42","first-page":"6397","article-title":"Scalable zero-shot entity linking with\n                        dense entity retrieval","volume-title":"Proceedings of the 2020\n                        Conference on Empirical Methods in Natural Language Processing\n                        (EMNLP)","author":"Ledell","year":"2020"},{"key":"2022033118512939700_bib43","doi-asserted-by":"publisher","first-page":"250","DOI":"10.18653\/v1\/K16-1025","article-title":"Joint learning of the embedding of words\n                        and entities for named entity disambiguation","volume-title":"Proceedings of The 20th SIGNLL Conference on Computational Natural\n                        Language Learning","author":"Yamada","year":"2016"},{"issue":"1","key":"2022033118512939700_bib44","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/1472-6947-15-S1-S4","article-title":"Entity linking for biomedical literature","volume":"15","author":"Zheng","year":"2015","journal-title":"BMC Medical Informatics and Decision Making"},{"key":"2022033118512939700_bib45","doi-asserted-by":"publisher","first-page":"243","DOI":"10.18653\/v1\/D19-6127","volume-title":"Towards zero-resource cross- lingual entity\n                        linking","author":"Zhou","year":"2019"}],"container-title":["Transactions of the Association for Computational Linguistics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/tacl\/article-pdf\/doi\/10.1162\/tacl_a_00460\/2004070\/tacl_a_00460.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/tacl\/article-pdf\/doi\/10.1162\/tacl_a_00460\/2004070\/tacl_a_00460.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,3,31]],"date-time":"2022-03-31T19:53:12Z","timestamp":1648756392000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/tacl\/article\/doi\/10.1162\/tacl_a_00460\/110051\/Multilingual-Autoregressive-Entity-Linking"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022]]},"references-count":45,"URL":"https:\/\/doi.org\/10.1162\/tacl_a_00460","relation":{},"ISSN":["2307-387X"],"issn-type":[{"value":"2307-387X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2022]]},"published":{"date-parts":[[2022]]}}}