5,657 Hits in 7.8 sec

Evaluation Benchmarks and Learning Criteria for Discourse-Aware Sentence Representations [article]

Mingda Chen, Zewei Chu, Kevin Gimpel
2019 arXiv   pre-print
We benchmark sentence encoders pretrained with our proposed training objectives, as well as other popular pretrained sentence encoders on DiscoEval and other sentence evaluation tasks.  ...  Prior work on pretrained sentence embeddings and benchmarks focus on the capabilities of stand-alone sentences.  ...  Acknowledgments We thank Jonathan Kummerfeld for helpful discussions about the IRC Disentanglement dataset, Davis Yoshida for discussions about BERT, and the anonymous reviewers for their feedback that  ... 
arXiv:1909.00142v2 fatcat:dylxptj4avgunlrmy337ztkrsu

Evaluation Benchmarks and Learning Criteria for Discourse-Aware Sentence Representations

Mingda Chen, Zewei Chu, Kevin Gimpel
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
Prior work on pretrained sentence embeddings and benchmarks focuses on the capabilities of representations for stand-alone sentences.  ...  We propose DiscoEval, a test suite of tasks to evaluate whether sentence representations include information about the role of a sentence in its discourse context.  ...  Acknowledgments We thank Jonathan Kummerfeld for helpful discussions about the IRC Disentanglement dataset, Davis Yoshida for discussions about BERT, and the anonymous reviewers for their feedback that  ... 
doi:10.18653/v1/d19-1060 dblp:conf/emnlp/ChenCG19 fatcat:tlpzcde6unfflo6ymuwrtbhjkq

Probing Multilingual Language Models for Discourse [article]

Murathan Kurfalı, Robert Östling
2021 arXiv   pre-print
We hope that our test suite, covering 5 tasks with a total of 22 languages in 10 distinct families, will serve as a useful evaluation platform for multilingual performance at and beyond the sentence level  ...  This is done with a systematic evaluation on a broader set of discourse-level tasks than has been previously been assembled.  ...  Evaluation benchmarks and learning criteria for discourse-aware sentence representations.  ... 
arXiv:2106.04832v1 fatcat:xmmmig243rguxmhfwi6t33aphy

Multilingual Neural RST Discourse Parsing [article]

Zhengyuan Liu, Ke Shi, Nancy F. Chen
2020 arXiv   pre-print
In this work, we investigate two approaches to establish a neural, cross-lingual discourse parser via: (1) utilizing multilingual vector representations; and (2) adopting segment-level translation of the  ...  Previous research under the Rhetorical Structure Theory (RST) has mostly focused on inducing and evaluating models from the English treebank.  ...  Webber for insightful discussions, C. Braud for sharing linguistic resources, and the anonymous reviewers for their precious feedback to help improve and extend this piece of work.  ... 
arXiv:2012.01704v1 fatcat:rfx4sqosife7fdc5zn7yaxz3qu

Automatic Story Generation: Challenges and Attempts [article]

Amal Alabdulkarim, Siyan Li, Xiangyu Peng
2021 arXiv   pre-print
Discuss future research directions and new technologies that may aid more advancements. 3. Shed light on emerging and often overlooked challenges such as creativity and discourse.  ...  Compared with COMeT, PARA-COMeT demonstrated the effectiveness of generating more implicit and novel discourse-aware inferences in paragraph level.  ...  Another example would be the paper by Gervás (2009) , where the author comments on storytelling systems and different evaluation criteria for creativity; similarly, all of the systems consist of planning  ... 
arXiv:2102.12634v1 fatcat:b67pi4zy5fc4dp4edidecwo54a

Conversational artificial intelligence - demystifying statistical vs linguistic NLP solutions

Kulvinder Panesar
2020 Journal of Computer-Assisted Linguistic Research  
Natural language is the most easily understood knowledge representation for people, but certainly not the best for computers because of its inherent ambiguous, complex and dynamic nature.  ...  Both are slowly emerging as a real presence in our lives from the impressive technological developments in machine learning, deep learning and natural language understanding solutions.  ...  Criteria 1 -Could the system present a mapping of the syntactic representation to a semantic representation, for a simple utterance?  ... 
doi:10.4995/jclr.2020.12932 fatcat:oogpuyd6zvhixi22k33xawe3dm

Discourse-Aware Unsupervised Summarization of Long Scientific Documents [article]

Yue Dong, Andrei Mircea, Jackie C. K. Cheung
2021 arXiv   pre-print
Our method assumes a two-level hierarchical graph representation of the source document, and exploits asymmetrical positional cues to determine sentence importance.  ...  Results on the PubMed and arXiv datasets show that our approach outperforms strong unsupervised baselines by wide margins in automatic metrics and human evaluation.  ...  We would like to thank Hao Zheng, Wen Xiao, and Sandeep Subramanian for useful discussions.  ... 
arXiv:2005.00513v2 fatcat:qncoatezxbantazwnlovy5n37a

Single Document Summarization as Tree Induction

Yang Liu, Ivan Titov, Mirella Lapata
2019 Proceedings of the 2019 Conference of the North  
Each root node in the tree is a summary sentence, and the subtrees attached to it are sentences whose content relates to or explains the summary sentence.  ...  We demonstrate experimentally on two benchmark datasets that our summarizer 1 performs competitively against state-of-the-art methods.  ...  Acknowledgments We thank Serhii Havrylov for helpful suggestions. This research is supported by a Google PhD Fellowship to the first author.  ... 
doi:10.18653/v1/n19-1173 dblp:conf/naacl/LiuTL19 fatcat:zs7yzicfsfhsnipskfwlixkuue

Enhancing Extractive Text Summarization with Topic-Aware Graph Neural Networks [article]

Peng Cui, Le Hu, Yuanchao Liu
2020 arXiv   pre-print
representation.  ...  Moreover, our model integrates a joint neural topic model (NTM) to discover latent topics, which can provide document-level features for sentence selection.  ...  We thank anonymous reviewers for their helpful comments on various aspects of this work.  ... 
arXiv:2010.06253v1 fatcat:kxlj4h2cszhcvh3uis5vlfs7ni

Scientific Statement Classification over [article]

Deyan Ginev, Bruce R. Miller
2019 arXiv   pre-print
We introduce a new classification task for scientific statements and release a large-scale dataset for supervised learning.  ...  Additionally, we introduce a lexeme serialization for mathematical formulas, and observe that context-aware models could improve when also trained on the symbolic modality.  ...  Llamapun implements its own word and sentence tokenization, aware of the formula modality.  ... 
arXiv:1908.10993v1 fatcat:i5rbegktgngt7f6xo2x64ln2iq

Exploring Input Representation Granularity for Generating Questions Satisfying Question-Answer Congruence

Madeeswaran Kannan, Haemanth Santhi Ponnusamy, Kordula De Kuthy, Lukas Stein, Detmar Meurers
2021 International Conference on Natural Language Generation  
In this paper, we explore the viability of form-based and more fine-grained encodings such as character or subword representations for question generation.  ...  Going beyond the specific task, this result adds to the evidence weighing different form-and meaning-based representations for natural language processing tasks.  ...  ., 2018) , and TyDi QA (Clark et al., 2020) are similarly unsuitable as they contain only little data, intended as benchmark for the evaluation of question answering systems.  ... 
dblp:conf/inlg/KannanPKSM21 fatcat:elvh7aqrv5abzdybkafhz6ea54

Discourse-Aware Neural Extractive Text Summarization [article]

Jiacheng Xu, Zhe Gan, Yu Cheng, Jingjing Liu
2020 arXiv   pre-print
DiscoBert extracts sub-sentential discourse units (instead of sentences) as candidates for extractive selection on a finer granularity.  ...  To address these issues, we present a discourse-aware neural summarization model - DiscoBert.  ...  Acknowledgement Thanks to Junyi Jessy Li, Greg Durrett, Yen-Chun Chen, and to the other members of the Microsoft Dynamics 365 AI Research team for the proofreading, feedback and suggestions.  ... 
arXiv:1910.14142v2 fatcat:qpl2e4o77rdlth4wh34x4wdvqm

Computational linguistics [article]

Antonio Toral Ruiz, Tommaso Caselli
2022 Zenodo  
The main theoretical and engineering goals are the understanding of natural language and the consequent development of tools and systems that can process and generate natural language, respectively.  ...  , semantics and pragmatics); (ii) an historical account of CL through the main computational approaches used in this field and the impact of these approaches on the relation between CL and other areas  ...  discourse representation (semantic level).  ... 
doi:10.5281/zenodo.6366253 fatcat:2dfs5gsfzzhlbmzaf4ow4jpuse

Formative Assessment Visual Feedback in Computer Graded Essays

Robert Francis Williams, Heinz Dreher
2005 Issues in Informing Science and Information Technology  
Concepts which are inadequately covered are displayed for the information of the author so that the essay can be improved.  ...  In this paper we discuss a simple but comprehensive form of feedback to essay authors, based on a thesaurus and computer graphics, which enables the essay authors to see where essay content is inadequate  ...  Discourse analysis techniques are used to examine the essay for discourse units by looking for surface cue words and non-lexical cues.  ... 
doi:10.28945/808 fatcat:7skqdmemknch3hfmantqfx6isa

Context-Preserving Text Simplification [article]

Christina Niklaus, Matthias Cetto, André Freitas, Siegfried Handschuh
2021 arXiv   pre-print
Using a set of linguistically principled transformation patterns, input sentences are converted into a hierarchical representation in the form of core sentences and accompanying contexts that are linked  ...  We present a context-preserving text simplification (TS) approach that recursively splits and rephrases complex English sentences into a semantic hierarchy of simplified sentences.  ...  Representa- tion learning for text-level discourse parsing.  ... 
arXiv:2105.11178v1 fatcat:c4feprhparfwhlr7pgptld5u54
« Previous Showing results 1 — 15 out of 5,657 results