检索结果-内蒙古大学图书馆

Fine-tuned neural models for Propaganda Detection at the sentence and fragment levels

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Alhindi, Tariq Pfeiffer, Jonas Muresan, Smaranda Department of Computer Science Columbia University Data Science Institute Columbia University Ubiquitous Knowledge Processing Lab Technische Universitat Darmstadt

This paper presents the CUNLP submission for the NLP4IF 2019 shared-task on Fine-Grained Propaganda Detection. Our system finished 5th out of 26 teams on the sentence-level classification task and 5th out of 11 teams on the fragment-level classification task based on our scores on the blind test set. We present our models, a discussion of our ablation studies and experiments, and an analysis of our performance on all eighteen propaganda techniques present in the corpus of the shared task. Copyright © 2019, The Authors. All rights reserved.

关键词： Classification (of information)

Does my rebuttal matter? Insights from a major NLP conference

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Gao, Yang Eger, Steffen Kuznetsov, Ilia Gurevych, Iryna Miyao, Yusuke Ubiquitous Knowledge Processing Lab Department of Computer Science Technische Universität Darmstadt Department of Computer Science Graduate School of Information Science and Technology University of Tokyo

Peer review is a core element of the scientific process, particularly in conference-centered fields such as ML and NLP. However, only few studies have evaluated its properties empirically. Aiming to fill this gap, we present a corpus that contains over 4k reviews and 1.2k author responses from ACL-2018. We quantitatively and qualitatively assess the corpus. This includes a pilot study on paper weaknesses given by reviewers and on quality of author responses. We then focus on the role of the rebuttal phase, and propose a novel task to predict after-rebuttal (i.e., final) scores from initial reviews and author responses. Although author responses do have a marginal (and statistically significant) influence on the final scores, especially for borderline papers, our results suggest that a reviewer's final score is largely determined by her initial score and the distance to the other reviewers' initial scores. In this context, we discuss the conformity bias inherent to peer reviewing, a bias that has largely been overlooked in previous research. We hope our analyses will help better assess the usefulness of the rebuttal phase in NLP conferences. Copyright © 2019, The Authors. All rights reserved.

关键词： Natural language processing systems

Revisiting joint modeling of cross-document entity and event coreference resolution

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Barhom, Shany Shwartz, Vered Eirew, Alon Bugert, Michael Reimers, Nils Dagan, Ido Computer Science Department Bar-Ilan University Intel AI Lab Israel Ubiquitous Knowledge Processing Lab Technische Universitat Darmstadt Germany

Recognizing coreferring events and entities across multiple texts is crucial for many NLP applications. Despite the task's importance, research focus was given mostly to withindocument entity coreference, with rather little attention to the other variants. We propose a neural architecture for cross-document coreference resolution. Inspired by Lee et al. (2012), we jointly model entity and event coreference. We represent an event (entity) mention using its lexical span, surrounding context, and relation to entity (event) mentions via predicate-arguments structures. Our model outperforms the previous state-of-the-art event coreference model on ECB+, while providing the first entity coreference results on this corpus. Our analysis confirms that all our representation elements, including the mention span itself, its context, and the relation to other mentions contribute to the model's success. Copyright © 2019, The Authors. All rights reserved.

关键词：

Better rewards yield better summaries: Learning to summarise without references

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Böhm, Florian Gao, Yang Meyer, Christian M. Shapira, Ori Dagan, Ido Gurevych, Iryna Ubiquitous Knowledge Processing Lab Technische Universität Darmstadt Germany Computer Science Department Bar-Ilan University Ramat-Gan Israel Dept. of Computer Science Royal Holloway University of London

Reinforcement Learning (RL) based document summarisation systems yield state-of-the-art performance in terms of ROUGE scores, because they directly use ROUGE as the rewards during training. However, summaries with high ROUGE scores often receive low human judgement. To find a better reward function that can guide RL to generate human-appealing summaries, we learn a reward function from human ratings on 2,500 summaries. Our reward function only takes the document and system summary as input. Hence, once trained, it can be used to train RL-based summarisation systems without using any reference summaries. We show that our learned rewards have significantly higher correlation with human ratings than previous approaches. Human evaluation experiments show that, compared to the state-of-the-art supervised-learning systems and ROUGE-as-rewards RL summarisation systems, the RL systems using our learned rewards during training generate summaries with higher human ratings. The learned reward function and our source code are available at https://***/yg211/ summary-reward-no-reference. Copyright © 2019, The Authors. All rights reserved.

关键词： Reinforcement learning

A Bayesian Approach for Sequence Tagging with Crowds

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Simpson, Edwin Gurevych, Iryna Department of Computer Science Ubiquitous Knowledge Processing Lab Technische Universität Darmstadt

Current methods for sequence tagging, a core task in NLP, are data hungry, which motivates the use of crowdsourcing as a cheap way to obtain labelled data. However, annotators are often unreliable and current aggregation methods cannot capture common types of span annotation errors. To address this, we propose a Bayesian method for aggregating sequence tags that reduces errors by modelling sequential dependencies between the annotations as well as the ground-truth labels. By taking a Bayesian approach, we account for uncertainty in the model due to both annotator errors and the lack of data for modelling annotators who complete few tasks. We evaluate our model on crowdsourced data for named entity recognition, information extraction and argument mining, showing that our sequential model outperforms the previous state of the art. We also find that our approach can reduce crowdsourcing costs through more effective active learning, as it better captures uncertainty in the sequence labels when there are few annotations. Copyright © 2018, The Authors. All rights reserved.

关键词： Crowdsourcing

How to Best Predict the Daily Number of New Infections of Covid-19

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Skiera, Bernd Jürgensmeier, Lukas Stowe, Kevin Gurevych, Iryna Board of EFL Data Science Institute Goethe University Frankfurt Theodor-W.-Adorno-Platz 4 Frankfurt60629 Germany Deakin University Australia Graduate School of Economics Finance and Management Goethe University Frankfurt Frankfurt60629 Germany Ubiquitous Processing Lab Computer Science Department Technical University of Darmstadt Hochschulstrasse 10 Darmstadt64289 Germany

knowledge about the daily number of new infections of Covid-19 is important because it is the basis for political decisions resulting in lockdowns and urgent health care measures. We use Germany as an example to illustrate shortcomings of official numbers, which are, at least in Germany, disclosed only with several days of delay and severely underreported on weekends (more than 40%). These shortcomings outline an urgent need for alternative data sources. The other widely cited source provided by the Center for Systems science and Engineering at Johns Hopkins University (JHU) also deviates for Germany on average by 79% from the official numbers. We argue that Google Search and Twitter data should complement official numbers. They predict even better than the original values from Johns Hopkins University and do so several days ahead. These two data sources could also be used in parts of the world where official numbers do not exist or are perceived to be unreliable. Copyright © 2020, The Authors. All rights reserved.

关键词：

A tool for extracting sense-disambiguated example sentences through user feedback 15

学校读者我要写书评

暂无评论

A tool for extracting sense-disambiguated example sentences ...

Software Demonstrations at the 15th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2017

作者： Boullosa, Beto De Castilho, Richard Eckart Geyken, Alexander Lemnitzer, Lothar Gurevych, Iryna Ubiquitous Knowledge Processing Lab Department of Computer Science Technische Universität Darmstadt Germany Berlin-Brandenburg Academy of Sciences Germany

ISBN: (纸本)9781510838604

This paper describes an application system aimed to help lexicographers in the extraction of example sentences for a given headword based on its different senses. The tool uses classification and clustering methods and incorporates user feedback to refine its results. © 2017 Association for Computational Linguistics.

关键词：

Corrigendum to Unsupervised Latent Dirichlet Allocation for supervised question classification. [Information processing & Management, 54(3), 380-393]

学校读者我要写书评

暂无评论

Information processing & Management 2019年第3期56卷 1080-1080页

作者： Saeedeh Momtazi Iryna Gurevych Department of Computer Engineering and Information Technology Amirkabir University of Technology Tehran Iran Ubiquitous Knowledge Processing (UKP) Lab TU Darmstadt

Assessing SRL frameworks with automatic training data expansion 11

学校读者我要写书评

暂无评论

Assessing SRL frameworks with automatic training data expans...

11th Linguistic Annotation Workshop, LAW 2017

作者： Hartmann, Silvana Mújdricza-Maydt, E. Éva Kuznetsov, Ilia Gurevych, Iryna Frank, Anette Ubiquitous Knowledge Processing Lab Department of Computer Science Technische Universität Darmstadt Germany Research Training School AIPHES Department of Computational Linguistics Heidelberg University Germany

ISBN: (纸本)9781945626395

We present the first experiment-based study that explicitly contrasts the three major semantic role labeling frameworks. As a prerequisite, we create a dataset labeled with parallel FrameNet-, PropBank-, and VerbNet-style labels for German. We train a state-of-the-art SRL tool for German for the different annotation styles and provide a comparative analysis across frameworks. We further explore the behavior of the frameworks with automatic training data generation. VerbNet provides larger semantic expressivity than PropBank, and we find that its generalization capacity approaches PropBank in SRL training, but it benefits less from training data expansion than the sparse-data affected FrameNet. © 2017 Association for Computational Linguistics.

关键词： Semantics