5,442 Hits in 8.5 sec

Linguistically-Informed Transformations (LIT): A Method for Automatically Generating Contrast Sets [article]

Chuanrong Li, Lin Shengshuo, Leo Z. Liu, Xinyi Wu, Xuhui Zhou, Shane Steinert-Threlkeld
2020 arXiv   pre-print
In this work, we propose a Linguistically-Informed Transformation (LIT) method to automatically generate contrast sets, which enables practitioners to explore linguistic phenomena of interests as well  ...  Experimenting with our method on SNLI and MNLI shows that current pretrained language models, although being claimed to contain sufficient linguistic knowledge, struggle on our automatically generated  ...  Acknowledgments We appreciate useful feedback from Noah A. Smith, Luke Zettlemoyer, Jungo Kasai, Yizhong Wang, and anonymous reviewers.  ... 
arXiv:2010.08580v3 fatcat:cmopfv47xnb4xggjiqfepddh4q

Linguistically-Informed Transformations (LIT): A Method for Automatically Generating Contrast Sets

Chuanrong Li, Lin Shengshuo, Zeyu Liu, Xinyi Wu, Xuhui Zhou, Shane Steinert-Threlkeld
2020 Proceedings of the Third BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP   unpublished
In this work, we propose a Linguistically-Informed Transformation (LIT) method to automatically generate contrast sets, which enables practitioners to explore linguistic phenomena of interests as well  ...  Experimenting with our method on SNLI and MNLI shows that current pretrained language models, although being claimed to contain sufficient linguistic knowledge, struggle on our automatically generated  ...  Generating Contrast Sets We propose a new Linguistically-Informed Transformation (LIT) method for large-scale automatic generation of contrast sets.  ... 
doi:10.18653/v1/2020.blackboxnlp-1.12 fatcat:zmustvjyqneh7chdzwizzbckse

"No Country for E-Lit?" – India and Electronic Literature

Souvik Mukherjee
2017 Hyperrhiz  
But the general explanation, offered by the Archaeological Survey of India and others, has been that these were just another set of caves set apart for meditation. 4 The elaborate carving in a series  ...  For certain keywords however -generally to do with race -Natural Selection would create a result set that linked to artist web sites Imaginations + Aesthetics / 203 DREAM DREAM about that keyword.  ...  Some of my programmes are locally generated but frequently I have to travel to find what I am looking for. I often feel that I am a hunter and gatherer.  ... 
doi:10.20415/hyp/016.e08 fatcat:ik7x2ivbxjdljoxfh4pyk3vt3y

Literal and Metaphorical Senses in Compositional Distributional Semantic Models

E.Dario Gutierrez, Ekaterina Shutova, Tyler Marghetis, Benjamin Bergen
2016 Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
We propose a method to learn metaphors as linear transformations in a vector space and find that, across a variety of semantic domains, explicitly modeling metaphor improves the resulting semantic representations  ...  Metaphorical expressions are pervasive in natural language and pose a substantial challenge for computational semantics.  ...  We then extend the generalizability of our approach by proposing a method to automatically learn metaphorical mappings as linear transformations in a CDSM.  ... 
doi:10.18653/v1/p16-1018 dblp:conf/acl/GutierrezSMB16 fatcat:rqbpv4kezfayznvxfv7nlvysce

In vitro evaluation of a program for machine-aided indexing

Christian Jacquemin, Béatrice Daille, Jean Royauté, Xavier Polanco
2002 Information Processing & Management  
We also thank the reviewers for their critics and suggestions, and Martine Hurault-Plantet (LIMSI) for her comments on a preliminary draft.  ...  Abstract This article presents the human evaluation of ILIAD, a program for Machine-Aided Indexing (MAI).  ...  MEDLARS is a system for accessing medical information that relies on human indexing methods and automatic search techniques.  ... 
doi:10.1016/s0306-4573(01)00050-4 fatcat:tyz5mivrwfgzzl26catu7p4zqi

Unsupervised Type and Token Identification of Idiomatic Expressions

Afsaneh Fazly, Paul Cook, Suzanne Stevenson
2009 Computational Linguistics  
In this article, we look into the usefulness of some of the identified linguistic properties of idioms for their automatic recognition.  ...  We use these statistical measures in a type-based classification task where we automatically separate idiomatic expressions (expressions with a possible idiomatic interpretation) from similar-on-the-surface  ...  Acknowledgments This article is an extended and updated combination of two papers that appeared, respectively, in the proceedings of EACL 2006 and the proceedings of the ACL 2007 Workshop on A Broader  ... 
doi:10.1162/coli.08-010-r1-07-048 fatcat:4xyn2iuk3zhdfntdj6t24cudum

Adaptive Attention Convolutional Neural Network for Liver Tumor Segmentation

Shunyao Luan, Xudong Xue, Yi Ding, Wei Wei, Benpeng Zhu
2021 Frontiers in Oncology  
For a large tumor, DG was 0.7819 and DC was 0.7632.ConclusionS-Net obtained more semantic information with the introduction of an attention mechanism and long jump connection.  ...  The segmentation DG for tumor was found to be 0.7555, DC was 0.613, VOE was 0.413, ASSD was 1.186 and RMSE was 1.804. For a small tumor, DG was 0.3246 and DC was 0.3082.  ...  ACKNOWLEDGMENTS We thank LetPub ( for its linguistic assistance during the preparation of this manuscript.  ... 
doi:10.3389/fonc.2021.680807 pmid:34434891 pmcid:PMC8381250 fatcat:rsvryvihmvebrclpvwfbvevaye

An English-Swahili parallel corpus and its use for neural machine translation in the news domain

Felipe Sánchez-Martínez, Víctor M. Sánchez-Cartagena, Juan Antonio Pérez-Ortiz, Mikel L. Forcada, Miquel Esplà-Gomis, Andrew Secker, Susie Coleman, Julie Wall
2020 Zenodo  
We report the results of a pilot human evaluation performed by the news media organisations participating in the H2020 EU-funded project GoURMET.  ...  This paper describes our approach to create a neural machine translation system to translate between English and Swahili (both directions) in the news domain, as well as the process we followed to crawl  ...  We thank the editors of the SAWA corpus for letting us use it for training. We also thank Wycliffe Muia (BBC) for help with Swahili examples and DW for helping in the manual evaluation.  ... 
doi:10.5281/zenodo.3923590 fatcat:lu5xrycnnfhqxeu6lzpqjthrlu

Effective Distant Supervision for Temporal Relation Extraction [article]

Xinyu Zhao, Shih-ting Lin, Greg Durrett
2021 arXiv   pre-print
We present a method of automatically collecting distantly-supervised examples of temporal relations.  ...  We demonstrate that a pre-trained Transformer model is able to transfer from the weakly labeled examples to human-annotated benchmarks in both zero-shot and few-shot settings, and that the masking scheme  ...  In this work, we present a method of automatically gathering distantly-labeled temporal relation examples.  ... 
arXiv:2010.12755v2 fatcat:omkd67k4kzgtpokljx2q6lki7y

Cerno: Light-weight tool support for semantic annotation of textual documents

Nadzeya Kiyavitskaya, Nicola Zeni, James R. Cordy, Luisa Mich, John Mylopoulos
2009 Data & Knowledge Engineering  
In this work, we present Cerno, a framework for semi-automatic semantic annotation of textual documents according to a domain-specific semantic model.  ...  These results suggest that light-weight semi-automatic techniques for semantic annotation are feasible, require limited human effort for adaptation to a new domain, and demonstrate markup quality comparable  ...  The tool doesn't need a set of seeds; instead it relies on automatically generated domainspecific extraction rules.  ... 
doi:10.1016/j.datak.2009.07.012 fatcat:u577vp6jwzf3noir3iqbthcblq

The IWSLT 2019 Evaluation Campaign

J. Niehues, R. Cattoni, S. Stüker, M. Negri, M. Turchi, T. Ha, E. Salesky, R. Sanabria, L. Barrault, L. Specia, M. Federico
2019 Zenodo  
For the first two tasks we encouraged submissions of end- to-end speech-to-text systems, and for the second task participants could also use the video as additional input.  ...  LIT LIT proposes layer-wise tied self-attention for end-to-end speech translation. Their method takes advantage of sharing weights of speech encoder and text decoder.  ...  In addition, for a more informative assessment automatic evaluation results were also computed in terms of case-insensitive BLEU, case-sensitive/insensitive TER [19] , BEER [20] , and CharacTER [21]  ... 
doi:10.5281/zenodo.3525577 fatcat:5gcfx7x3bfbe3gea3g6ocoyp4e

Thermostat: A Large Collection of NLP Model Explanations and Analysis Tools [article]

Nils Feldhus, Robert Schwarzenberg, Sebastian Möller
2021 arXiv   pre-print
Thermostat allows easy access to over 200k explanations for the decisions of prominent state-of-the-art models spanning across different NLP tasks, generated with multiple explainers.  ...  To facilitate research, we present Thermostat which consists of a large collection of model explanations and accompanying analysis tools.  ...  Acknowledgements We would like to thank Lisa Raithel, Steffen Castle, and David Harbecke for their valuable feedback.  ... 
arXiv:2108.13961v1 fatcat:gazrqwyaj5eylhosttogwue7yu

Empirical Linguistic Study of Sentence Embeddings

Katarzyna Krasnowska-Kieraś, Alina Wróblewska
2019 Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics  
We introduce a method of analysing the content of sentence embeddings based on universal probing tasks, along with the classification datasets for two contrasting languages.  ...  The purpose of the research is to answer the question whether linguistic information is retained in vector representations of sentences.  ...  In the probing-based scenario, a set of language-independent tests was designed and probing datasets were generated for two contrasting languages -English and Polish.  ... 
doi:10.18653/v1/p19-1573 dblp:conf/acl/Krasnowska-Kieras19 fatcat:5kwxe5f3urfehhmlk7u6d3tbpe

Induction of Dependency Structures Based on Weighted Projection [chapter]

Alina Wróblewska, Adam Przepiórkowski
2012 Lecture Notes in Computer Science  
This paper describes a novel weighted projection method of inducing grammatical dependency structures for Polish.  ...  Minimum spanning trees induced from such graphs are used to train a parsing model with a publicly available parser-generation system.  ...  method of cross-lingual projection of linguistic information.  ... 
doi:10.1007/978-3-642-34630-9_38 fatcat:ikwe77f3wja6dl4yw4afoq5o7q

Automatic Acquisition and Expansion of Hypernym Links

Emmanuel Morin, Christian Jacquemin
2004 Language Resources and Evaluation  
This paper proposes to bridge the gap between term acquisition and thesaurus construction by offering a framework for automatic structuring of multi-word candidate terms with the help of corpus-based links  ...  The induced hierarchy is incomplete but provides an automatic generalization of singleword terms relations to multi-word terms that are pervasive in technical thesauri and corpora.  ...  A common environment relative to a set of sentences is extracted automatically by the previous method and manually by Hearst. 2.  ... 
doi:10.1007/s10579-004-1926-2 fatcat:7jmqws745fhutglojqga2a7r3q
« Previous Showing results 1 — 15 out of 5,442 results