Filters








2,800 Hits in 9.5 sec

Toward Finding Semantic Relations not Written in a Single Sentence: An Inference Method using Auto-Discovered Rules

Masaaki Tsuchida, Kentaro Torisawa, Stijn De Saeger, Jong-Hoon Oh, Jun'ichi Kazama, Chikara Hashimoto, Hayato Ohwada
2011 International Joint Conference on Natural Language Processing  
We develop a method to infer new semantic relation instances by applying auto-discovered inference rules, and show that our method inferred a considerable number of valid instances that were not written  ...  Our goal in this work is to obtain valid non-single sentence relation instances, which are not written in any single sentence and may not be even written in a large corpus.  ...  NS instances may be indirectly written in a sequence of several sentences using anaphora or may not be written at all even in a given corpus.  ... 
dblp:conf/ijcnlp/TsuchidaTSOKHO11 fatcat:wj3wxz3wh5b7bphuozdm7fdd6y

Unsupervised Sentence Compression using Denoising Auto-Encoders [article]

Thibault Févry, Jason Phang
2018 arXiv   pre-print
To remove the need for paired corpora, we emulate a summarization task and add noise to extend sentences and train a denoising auto-encoder to recover the original, constructing an end-to-end training  ...  In sentence compression, the task of shortening sentences while retaining the original meaning, models tend to be trained on large corpora containing pairs of verbose and compressed sentences.  ...  Acknowledgments We would like to express our gratitude to Sam Bowman for his thoughtful advice and feedback in the writing of this paper. We thank the NVIDIA Corporation for their support.  ... 
arXiv:1809.02669v1 fatcat:dmxfnlwucfbxtabohyq5m5at5e

Unsupervised Sentence Compression using Denoising Auto-Encoders

Thibault Fevry, Jason Phang
2018 Proceedings of the 22nd Conference on Computational Natural Language Learning  
To remove the need for paired corpora, we emulate a summarization task and add noise to extend sentences and train a denoising auto-encoder to recover the original, constructing an end-to-end training  ...  In sentence compression, the task of shortening sentences while retaining the original meaning, models tend to be trained on large corpora containing pairs of verbose and compressed sentences.  ...  Acknowledgments We would like to express our deepest gratitude to Sam Bowman for his thoughtful advice and feedback in the writing of this paper. We thank the NVIDIA Corporation for their support.  ... 
doi:10.18653/v1/k18-1040 dblp:conf/conll/FevryP18 fatcat:3247tqigbfgjlgsoybdyyp3o3m

Extending Multi-Sense Word Embedding to Phrases and Sentences for Unsupervised Semantic Applications [article]

Haw-Shiuan Chang, Amol Agrawal, Andrew McCallum
2021 arXiv   pre-print
In phrase similarity experiments, we discover that the multi-facet embeddings provide an interpretable semantic representation but do not outperform the single-facet baseline.  ...  We propose a novel embedding method for a text sequence (a phrase or a sentence) where each sequence is represented by a distinct set of multi-mode codebook embeddings to capture different semantic facets  ...  Thus, we recommend adopting our method when the tion methods that do not use the sentence order information source of training and testing sentences are not too different versus the sentence  ... 
arXiv:2103.15330v2 fatcat:qv7rz7kl2ndovfrgcra2lixz6y

Resurrection from a Death Sentence: Why Capital Sentences Should Be Commuted upon the Occasion of an Authentic Ethical Transformation

B. Douglas Robbins
2001 University of Pennsylvania law review  
ISSUEs 19, 27, 32 (1994) (finding that people's attitudes toward the death penalty are not determined by their beliefs in its deterrent effectiveness); Tyler & Weber, supra, at 26 (finding that sixty-six  ...  which a scientist's personal values enter into the scientific process: First, scientific facts are not immanent in an objective reality waiting to be discovered by any scientists who look in the right  ...  Murray, 476 U.S. 28, 36-37 (1986) (invalidating a death sentence as violative of the Sixth Amendment right to an impartial jury where a black de-  ... 
doi:10.2307/3312991 fatcat:nfwitlaykbhsvcgbs6luecnja4

5-2 音声質問応答システム「一休」

DE SAEGER Stijn, VARGA István
2012 Journal of NICT  
able to generate answer hypotheses that are not written explicitly but can be derived by combining seemingly unrelated information obtained from distinct documents.  ...  Ikkyu is a next-generation information system that caters to users' various information needs, in the form of spoken natural language questions posed via smartphone.  ...  Finding Semantic Relations not Written in a Single Sentence: An Inference Method using Auto-Discovered Rules, In Proceedings of the 5th International Joint Conference on Natural Language Processing (IJCNLP  ... 
doi:10.24812/nictkenkyuhoukoku.58.3.4_77 fatcat:s2cjwifmcnfr5luaa7lmdv7ffu

Extractive Summarization as Text Matching [article]

Ming Zhong, Pengfei Liu, Yiran Chen, Danqing Wang, Xipeng Qiu, Xuanjing Huang
2020 arXiv   pre-print
Instead of following the commonly used framework of extracting sentences individually and modeling the relationship between sentences, we formulate the extractive summarization task as a semantic text  ...  matching problem, in which a source document and candidate summaries will be (extracted from the original text) matched in a semantic space.  ...  We find that there is indeed an inherent gap between the two approaches across these datasets, which motivates us to propose the following summary-level method.  ... 
arXiv:2004.08795v1 fatcat:hjzataeqfzhglafxkc7vxydrte

Sentencing the Green-Collar Offender: Punishment, Culpability, and Environmental Crime

Michael M. O'Hear
2004 Journal of criminal law & criminology  
Allenbaugh, The Reason Behind the Rules: Finding and Using the Philosophy of the Federal Sentencing Guidelines, 40 AM. CRIM. L.  ...  In this view, judges are trying to use the departure mechanism to correct the tendency of the guidelines towards disproportionate sentences for low-culpability offenders. 40 7 This view finds additional  ...  the piecemeal approach, it is better to have a system that displays trust in judges and accords them discretion at sentencing in an open and honest manner.  ... 
doi:10.2307/3491383 fatcat:iw4qosgx55ewzhgjypvch5qdeq

Learning Interpretable and Discrete Representations with Adversarial Training for Unsupervised Text Classification [article]

Yau-Shian Wang and Hung-Yi Lee and Yun-Nung Chen
2020 arXiv   pre-print
Also, the performance is on par with a recently proposed weakly-supervised text classification method.  ...  The discrete code learned by TIGAN can be used for unsupervised text classification.  ...  Most of the topical words in a same topic are not semantically related to each other.  ... 
arXiv:2004.13255v1 fatcat:jodrqpdmeba73d3odvknooqami

Analyzing and Interpreting Neural Networks for NLP: A Report on the First BlackboxNLP Workshop [article]

Afra Alishahi and Grzegorz Chrupała and Tal Linzen
2019 arXiv   pre-print
Here we review a number of representative studies in each category.  ...  They use this method to examine how representations evolve over time and find that POS tag information is acquired before semantic or topics information.  ...  English sentence, a parser based on a precision grammar can be used to determine if a sentence is grammatical in English.  ... 
arXiv:1904.04063v1 fatcat:iwuio4l62jcpjindqpyzs6rmdy

Call by Meaning

Hesam Samimi, Chris Deaton, Yoshiki Ohshima, Alessandro Warth, Todd Millstein
2014 Proceedings of the 2014 ACM International Symposium on New Ideas, New Paradigms, and Reflections on Programming & Software - Onward! '14  
Inspired by the aspirations of SW, in this paper we reformulate its visions from the perspective of a programming model, i.e., that components themselves should be able to interact using semantic ontologies  ...  A multitude of solutions have been proposed under the umbrella term Semantic Web (SW), in which semantic markup of the components with concepts from semantic ontologies and the ability to invoke queries  ...  Acknowledgments We thank Alan Kay for the encouragement to explore Cyc towards programming.  ... 
doi:10.1145/2661136.2661152 dblp:conf/oopsla/SamimiDOWM14 fatcat:uj7qmal5mzgkhhcxmcinna2jnm

Engineering the Production of Meta-Information: The Abstracting Concern

Maria Pinto
2003 Journal of information science  
In addition, reticular and graphic models of information representation, much more appropriate to digital environments, offer a complementary method.  ...  However, the low quality of the product derived from full automation (extract and summaries), above all lacking in coherence, led us to the concept of partial automation, a hybrid man-machine methodology  ...  If not, the selected sentence is re-written so that it makes sense on its own.  ... 
doi:10.1177/01655515030295006 fatcat:li77lizccjh4hm43ammkqchgya

Semi-Automated Protocol Disambiguation and Code Generation [article]

Jane Yen and Tamás Lévai and Qinyuan Ye and Xiang Ren and Ramesh Govindan and Barath Raghavan
2021 arXiv   pre-print
For decades, Internet protocols have been specified using natural language. Given the ambiguity inherent in such text, it is not surprising that protocol implementations have long exhibited bugs.  ...  Using SAGE, we discover 5 instances of ambiguity and 6 instances of under-specification in the ICMP RFC; after clarification, SAGE is able to automatically generate code that interoperates perfectly with  ...  It also discovered 6 sentences that failed unit tests (all of which are variants of a single sentence).  ... 
arXiv:2010.04801v2 fatcat:p6gc6mzknjctrbexu2zfoykgpm

Message from the general chair

Benjamin C. Lee
2015 2015 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)  
methods as an isolated inference procedure at the end.  ...  To inject knowledge, we use a state-of-the-art system which cross-links (or "grounds") expressions in free text to Wikipedia.  ...  Previous sentence similarity work finds that latent semantics approaches to the problem do not perform well due to insufficient information in single sentences.  ... 
doi:10.1109/ispass.2015.7095776 dblp:conf/ispass/Lee15 fatcat:ehbed6nl6barfgs6pzwcvwxria

Design considerations for a hierarchical semantic compositional framework for medical natural language understanding [article]

Ricky K. Taira, Anders O. Garlid, William Speier
2022 arXiv   pre-print
We discuss the design of a generative semantic model and an associated semantic parser used to transform a free-text sentence into a logical representation of its meaning.  ...  In this paper, we describe a framework inspired by mechanisms of human cognition in an attempt to jump the NLP performance curve.  ...  We would also like to thank Lew Andrada for grammar-related edits.  ... 
arXiv:2204.02067v1 fatcat:k7s5sirbw5gs7p7oi4xha5m5gq
« Previous Showing results 1 — 15 out of 2,800 results