Filters








80 Hits in 3.3 sec

A BiLSTM-based System for Cross-lingual Pronoun Prediction

Sara Stymne, Sharid Loáiciga, Fabienne Cap
<span title="">2017</span> <i title="Association for Computational Linguistics"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/wiyp6rhsjbhrzl3gyyy6u5xcj4" style="color: black;">Proceedings of the Third Workshop on Discourse in Machine Translation</a> </i> &nbsp;
The system is based on a lower layer of BiLSTMs reading the source and target sentences respectively.  ...  We describe the Uppsala system for the 2017 DiscoMT shared task on crosslingual pronoun prediction.  ...  FC was funded by a VINNMER Marie Curie Incoming Grant within VINNO-VAs Mobility for Growth programme. Computations were completed in the Taito-CSC cluster in Helsinki through NeIC-NLPL (www.nlpl.eu).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/w17-4805">doi:10.18653/v1/w17-4805</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/discomt/StymneLC17.html">dblp:conf/discomt/StymneLC17</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/v6qlj4ytvfan5ke73qh6nqxzd4">fatcat:v6qlj4ytvfan5ke73qh6nqxzd4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200309100901/https://www.aclweb.org/anthology/W17-4805.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/27/ba/27ba95b576567bdaf568d877e3fafc2233190ba1.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/w17-4805"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Findings of the 2017 DiscoMT Shared Task on Cross-lingual Pronoun Prediction

Sharid Loáiciga, Sara Stymne, Preslav Nakov, Christian Hardmeier, Jörg Tiedemann, Mauro Cettolo, Yannick Versley
<span title="">2017</span> <i title="Association for Computational Linguistics"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/wiyp6rhsjbhrzl3gyyy6u5xcj4" style="color: black;">Proceedings of the Third Workshop on Discourse in Machine Translation</a> </i> &nbsp;
We describe the design, the setup, and the evaluation results of the DiscoMT 2017 shared task on cross-lingual pronoun prediction.  ...  The task asked participants to predict a target-language pronoun given a source-language pronoun in the context of a sentence.  ...  We thank Andrei Popescu-Belis and Bonnie Webber for their advice in organizing this shared task.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/w17-4801">doi:10.18653/v1/w17-4801</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/discomt/LoaicigaSNHTCV17.html">dblp:conf/discomt/LoaicigaSNHTCV17</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/sjwcdgwrlvaovp2nls5zeht73a">fatcat:sjwcdgwrlvaovp2nls5zeht73a</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20180719085737/http://uu.diva-portal.org/smash/get/diva2:1172487/FULLTEXT01" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/14/84/1484081d165c7566a19ea9b52bbab0fd5bfa699c.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/w17-4801"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

A systematic comparison of methods for low-resource dependency parsing on genuinely low-resource languages [article]

Clara Vania, Yova Kementchedjhieva, Anders Søgaard, Adam Lopez
<span title="2019-09-06">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We systematically compare a set of simple strategies for improving low-resource parsers: data augmentation, which has not been tested before; cross-lingual training; and transliteration.  ...  related high-resource treebank is available, cross-lingual training is helpful and complements data augmentation; and (3) when the high-resource treebank uses a different writing system, transliteration  ...  Anders Søgaard is supported by a Google Focused Research Award. We thank Aibek Makazhanov for helping with Kazakh transliteration, and Miryam de Lhoneux for parser implementation.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1909.02857v1">arXiv:1909.02857v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/uuraw5arvvg3bpyje4s3eqbdey">fatcat:uuraw5arvvg3bpyje4s3eqbdey</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200903021615/https://arxiv.org/pdf/1909.02857v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/55/ac/55ace73cb8f7c3c4bd6d8ead45c0ba6193d1afda.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1909.02857v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

A systematic comparison of methods for low-resource dependency parsing on genuinely low-resource languages

Clara Vania, Yova Kementchedjhieva, Anders Søgaard, Adam Lopez
<span title="">2019</span> <i title="Association for Computational Linguistics"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/u3ideoxy4fghvbsstiknuweth4" style="color: black;">Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)</a> </i> &nbsp;
related high-resource treebank is available, cross-lingual training is helpful and complements data augmentation; and (3) when the high-resource treebank uses a different writing system, transliteration  ...  We systematically compare a set of simple strategies for improving low-resource parsers: data augmentation, which has not been tested before; cross-lingual training; and transliteration.  ...  Anders Søgaard is supported by a Google Focused Research Award. We thank Aibek Makazhanov for helping with Kazakh transliteration, and Miryam de Lhoneux for parser implementation.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/d19-1102">doi:10.18653/v1/d19-1102</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/emnlp/VaniaKSL19.html">dblp:conf/emnlp/VaniaKSL19</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/w6tt4bygofefxbtgdepnoktfku">fatcat:w6tt4bygofefxbtgdepnoktfku</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201103103105/https://curis.ku.dk/ws/files/240407657/OA_A_systematic_comparison_of_methods_for_low_resource_dependency_parsing.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/b9/2e/b92ee9ed4cf4a55560d4fa5d77ae4f3bee17fb35.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/d19-1102"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Learning when to trust distant supervision: An application to low-resource POS tagging using cross-lingual projection

Meng Fang, Trevor Cohn
<span title="">2016</span> <i title="Association for Computational Linguistics"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/ot5sbt27tzdyrhcooo2wxlw7ki" style="color: black;">Proceedings of The 20th SIGNLL Conference on Computational Natural Language Learning</a> </i> &nbsp;
In this paper, we introduce a novel approach to sequence tagging that learns to correct the errors from cross-lingual projection using an explicit debiasing layer.  ...  Cross lingual projection of linguistic annotation suffers from many sources of bias and noise, leading to unreliable annotations that cannot be used directly.  ...  Although cross-lingual POS projection is popular it has several problems, including errors from poor word alignments and cross-lingual syntactic divergence .  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/k16-1018">doi:10.18653/v1/k16-1018</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/conll/FangC16.html">dblp:conf/conll/FangC16</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/tmsnbysb2zej5fo22idvi4pmoa">fatcat:tmsnbysb2zej5fo22idvi4pmoa</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200507203259/https://www.aclweb.org/anthology/K16-1018.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/54/1d/541d67f6ae026af328a5d46483809f509ec9b3ed.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/k16-1018"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Learning when to trust distant supervision: An application to low-resource POS tagging using cross-lingual projection [article]

Meng Fang, Trevor Cohn
<span title="2016-07-05">2016</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this paper, we introduce a novel approach to sequence tagging that learns to correct the errors from cross-lingual projection using an explicit debiasing layer.  ...  Cross lingual projection of linguistic annotation suffers from many sources of bias and noise, leading to unreliable annotations that cannot be used directly.  ...  Although cross-lingual POS projection is popular it has several problems, including errors from poor word alignments and cross-lingual syntactic divergence .  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1607.01133v1">arXiv:1607.01133v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/dwz6vcqxi5extb5te42evdl2ly">fatcat:dwz6vcqxi5extb5te42evdl2ly</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200907040325/https://arxiv.org/pdf/1607.01133v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ab/4e/ab4e416dd69970cc35ee2c4808bcccfc3756d1c8.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1607.01133v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Evaluating Pronominal Anaphora in Machine Translation: An Evaluation Measure and a Test Suite [article]

Prathyusha Jwalapuram, Shafiq Joty, Irina Temnikova, Preslav Nakov
<span title="2019-08-31">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
real system translations, for English.  ...  With this aim in mind, we contribute an extensive, targeted dataset that can be used as a test suite for pronoun translation, covering multiple source languages and different pronoun errors drawn from  ...  Acknowledgments We would like to thank the anonymous reviewers for their comments. Shafiq Joty would like to thank the funding support from his Start-up Grant (M4082038.020).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1909.00131v1">arXiv:1909.00131v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/y3lgckrn2nfsney2wxix5omaxu">fatcat:y3lgckrn2nfsney2wxix5omaxu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200826204123/https://arxiv.org/pdf/1909.00131v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/85/2c/852cad84715768db14912303771949e0488597a2.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1909.00131v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Evaluating Pronominal Anaphora in Machine Translation: An Evaluation Measure and a Test Suite

Prathyusha Jwalapuram, Shafiq Joty, Irina Temnikova, Preslav Nakov
<span title="">2019</span> <i title="Association for Computational Linguistics"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/u3ideoxy4fghvbsstiknuweth4" style="color: black;">Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)</a> </i> &nbsp;
real system translations, for English.  ...  With this aim in mind, we contribute an extensive, targeted dataset that can be used as a test suite for pronoun translation, covering multiple source languages and different pronoun errors drawn from  ...  Acknowledgments We would like to thank the anonymous reviewers for their comments. Shafiq Joty would also like to thank the funding support from his Start-up Grant (M4082038.020).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/d19-1294">doi:10.18653/v1/d19-1294</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/emnlp/JwalapuramJTN19.html">dblp:conf/emnlp/JwalapuramJTN19</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/apjb37qq2venfkqtblirydm3xy">fatcat:apjb37qq2venfkqtblirydm3xy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200117120153/https://www.aclweb.org/anthology/D19-1294.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/78/d8/78d8ea36ad670c9b9bf7f8a8b42f9e4a66edf6ef.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/d19-1294"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Joint learning of morphology and syntax with cross-level contextual information flow

Burcu Can, Hüseyin Aleçakır, Suresh Manandhar, Cem Bozşahin
<span title="2022-01-20">2022</span> <i title="Cambridge University Press (CUP)"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/ugfrgjxxffftblvvdh5ko6pyum" style="color: black;">Natural Language Engineering</a> </i> &nbsp;
We propose an integrated deep learning model for morphological segmentation, morpheme tagging, part-of-speech (POS) tagging, and syntactic parsing onto dependencies, using cross-level contextual information  ...  Primary focus is agglutination in morphology, in particular Turkish morphology, for which we demonstrate improved performance compared to models trained for individual tasks.  ...  Based on the predicted outputs for each position inside the word, we compute the loss defined by binary cross-entropy as follows for Y number of characters inside a word: L seg = − Y i=1 y log( ŷi ) −  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1017/s1351324921000371">doi:10.1017/s1351324921000371</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/e4n6xt5uwjc4ni5femc765zesm">fatcat:e4n6xt5uwjc4ni5femc765zesm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220201141156/https://www.cambridge.org/core/services/aop-cambridge-core/content/view/6D848C753AAA7DD978217645283F9DFE/S1351324921000371a.pdf/div-class-title-joint-learning-of-morphology-and-syntax-with-cross-level-contextual-information-flow-div.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/a0/b3/a0b32c03be7f7eb89bcee9bd699c16c4a9ccfc38.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1017/s1351324921000371"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> cambridge.org </button> </a>

Zero-resource Multi-dialectal Arabic Natural Language Understanding

Muhammad Khalifa, Hesham Hassan, Aly Fahmy
<span title="">2021</span> <i title="The Science and Information Organization"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/2yzw5hsmlfa6bkafwsibbudu64" style="color: black;">International Journal of Advanced Computer Science and Applications</a> </i> &nbsp;
A reasonable amount of annotated data is required for fine-tuning pre-trained language models (PLM) on downstream tasks.  ...  In this paper, we investigate the zero-shot performance on Dialectal Arabic (DA) when fine-tuning a PLM on modern standard Arabic (MSA) data only – identifying a significant performance drop when evaluating  ...  Baselines For the NER task, we use the following baselines: • NERA [31]: A hybrid system of rule-based features and a decision tree classifier. • WC-BiLSTM [21] : A character-and a word-level Bi-LSTM  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.14569/ijacsa.2021.0120369">doi:10.14569/ijacsa.2021.0120369</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/kw57zdjmmbcjzg42gaa2372bnu">fatcat:kw57zdjmmbcjzg42gaa2372bnu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210403055156/https://thesai.org/Downloads/Volume12No3/Paper_69-Zero_resource_Multi_dialectal_Arabic_Natural_Language.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ea/d5/ead5076b02ca1358d780b901c86c69dfe751653e.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.14569/ijacsa.2021.0120369"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> Publisher / doi.org </button> </a>

Parser Training with Heterogeneous Treebanks [article]

Sara Stymne, Miryam de Lhoneux, Aaron Smith, Joakim Nivre
<span title="2018-05-14">2018</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We go on to propose a new method based on treebank embeddings.  ...  We start by investigating previously suggested, but little evaluated, strategies for exploiting multiple treebanks based on concatenating training sets, with or without fine-tuning.  ...  Unlike much work on cross-lingual parsing, we do not focus on a low-resource scenario.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1805.05089v1">arXiv:1805.05089v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/az3m3257cjh2bchzpjzrzeuyfe">fatcat:az3m3257cjh2bchzpjzrzeuyfe</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200826043317/https://arxiv.org/pdf/1805.05089v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/68/76/68769d95a114c059f8686af224655f8f246ec3df.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1805.05089v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

What does it mean to be language-agnostic? Probing multilingual sentence encoders for typological properties [article]

Rochelle Choenni, Ekaterina Shutova
<span title="2020-09-27">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Multilingual sentence encoders have seen much success in cross-lingual model transfer for downstream NLP tasks.  ...  We propose methods for probing sentence representations from state-of-the-art multilingual encoders (LASER, M-BERT, XLM and XLM-R) with respect to a range of typological properties pertaining to lexical  ...  Multilingual sentence encoders LASER is a BiLSTM encoder trained with an encoder-decoder architecture and a cross-lingual objective -machine translation (MT).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2009.12862v1">arXiv:2009.12862v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ev3eo6zu2zhjfd6gwst4weizkm">fatcat:ev3eo6zu2zhjfd6gwst4weizkm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201014181051/https://arxiv.org/pdf/2009.12862v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2009.12862v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

HIT: A Hierarchically Fused Deep Attention Network for Robust Code-mixed Language Representation [article]

Ayan Sengupta, Sourabh Kumar Bhattacharjee, Tanmoy Chakraborty, Md Shad Akhtar
<span title="2021-05-30">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
HIT is a hierarchical transformer-based framework that captures the semantic relationship among words and hierarchically learns the sentence-level semantics using a fused attention mechanism.  ...  In this paper, we propose HIT, a robust representation learning method for code-mixed texts.  ...  Acknowledgement The work was partially supported by the Ramanujan Fellowship (SERB) and the Infosys Centre for AI, IIITD.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2105.14600v1">arXiv:2105.14600v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/bqtnemsqmvb5zitwnz6rjdl3n4">fatcat:bqtnemsqmvb5zitwnz6rjdl3n4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210607064518/https://arxiv.org/pdf/2105.14600v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/94/08/940824fb28d4c92382e481e8250100925ef83c87.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2105.14600v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Self-Training Pre-Trained Language Models for Zero- and Few-Shot Multi-Dialectal Arabic Sequence Labeling [article]

Muhammad Khalifa and Muhammad Abdul-Mageed and Khaled Shaalan
<span title="2021-02-02">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
A sufficient amount of annotated data is usually required to fine-tune pre-trained language models for downstream tasks.  ...  We demonstrate the utility of our approach in the context of Arabic sequence labeling by using a language model fine-tuned on Modern Standard Arabic (MSA) only to predict named entities (NE) and part-of-speech  ...  Baselines For the NER task, we use the following baselines: • NERA (Abdallah et al., 2012) : A hybrid system of rule-based features and a decision tree classifier. • WC-BiLSTM (Gridach, 2016) : A characterand  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2101.04758v4">arXiv:2101.04758v4</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/lnoy3lrstjezdflu6klsh52s7u">fatcat:lnoy3lrstjezdflu6klsh52s7u</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210213082621/https://arxiv.org/pdf/2101.04758v4.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/7f/1f/7f1f473e6042275149d3589022b76232e3c4c2f5.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2101.04758v4" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Parser Training with Heterogeneous Treebanks

Sara Stymne, Miryam de Lhoneux, Aaron Smith, Joakim Nivre
<span title="">2018</span> <i title="Association for Computational Linguistics"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/5n6volmnonf5tn6xputi5f2t3e" style="color: black;">Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)</a> </i> &nbsp;
We go on to propose a new method based on treebank embeddings.  ...  We start by investigating previously suggested, but little evaluated, strategies for exploiting multiple treebanks based on concatenating training sets, with or without fine-tuning.  ...  Unlike much work on cross-lingual parsing, we do not focus on a low-resource scenario.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/p18-2098">doi:10.18653/v1/p18-2098</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/acl/StymneLSN18.html">dblp:conf/acl/StymneLSN18</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/i5d5obx7a5hxxh4xhtolbb2dzm">fatcat:i5d5obx7a5hxxh4xhtolbb2dzm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200507182341/https://www.aclweb.org/anthology/P18-2098.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/02/b7/02b7c4a2600fd9e6c0860b92a02ff2cce8910abe.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/p18-2098"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>
&laquo; Previous Showing results 1 &mdash; 15 out of 80 results