Results 271 to 280 of about 1,557,530 (322)

Revisiting Zero-Shot Abstractive Summarization in the Era of Large Language Models from the Perspective of Position Bias

North American Chapter of the Association for Computational Linguistics
We characterize and study zero-shot abstractive summarization in Large Language Models (LLMs) by measuring position bias, which we propose as a general formulation of the more restrictive lead bias phenomenon studied previously in the literature ...
Anshuman Chhabra, Hadi Askari, Prasant Mohapatra   +2 more
semanticscholar   +1 more source

Zero-subject Resolution Using Linguistic Constraints and Defaults: The Case of Japanese Instruction Manuals

Machine Translation, 1999
This paper proposes a method for anaphora resolution of zero subjects in Japanese instruction manuals based on both the linguistic nature of expressions and the general ontology of the text type. In instruction manuals written in Japanese, zero subject is one of main reasons for ambiguity of sentences. In order to resolve them, we examined the property
Tatsunori Mori, Mamoru Matsuo, Hiroshi Nakawaga   +2 more
openaire   +1 more source

Zero-shot Sentiment Analysis in Low-Resource Languages Using a Multilingual Sentiment Lexicon

Conference of the European Chapter of the Association for Computational Linguistics
Improving multilingual language models capabilities in low-resource languages is generally difficult due to the scarcity of large-scale data in those languages. In this paper, we relax the reliance on texts in low-resource languages by using multilingual
Fajri Koto, T. Beck, Zeerak Talat, Iryna Gurevych, Timothy Baldwin   +4 more
semanticscholar   +1 more source

Smaller Language Models are Better Zero-shot Machine-Generated Text Detectors

Conference of the European Chapter of the Association for Computational Linguistics
As large language models are becoming more embedded in different user-facing services, it is important to be able to distinguish between human-written and machine-generated text to verify the authenticity of news articles, product reviews, etc.
Niloofar Mireshghallah, Justus Mattern, Sicun Gao, Reza Shokri, Taylor Berg-Kirkpatrick   +4 more
semanticscholar   +1 more source

LLM-Driven Knowledge Injection Advances Zero-Shot and Cross-Target Stance Detection

North American Chapter of the Association for Computational Linguistics
Stance detection aims at inferring an author’s attitude towards a specific target in a text. Prior methods mainly consider target-related background information for a better understanding of targets while neglecting the accompanying input texts.
Zhao Zhang, Yiming Li, Jin Zhang, Hui Xu
semanticscholar   +1 more source

Analysing Zero-Shot Readability-Controlled Sentence Simplification

International Conference on Computational Linguistics
Readability-controlled text simplification (RCTS) rewrites texts to lower readability levels while preserving their meaning. RCTS models often depend on parallel corpora with readability annotations on both source and target sides.
Abdullah Barayan, José Camacho-Collados, Fernando Alva-Manchego   +2 more
semanticscholar   +1 more source

Tree-of-Counterfactual Prompting for Zero-Shot Stance Detection

Annual Meeting of the Association for Computational Linguistics
Stance detection enables the inference of atti-001 tudes from human communications. Automatic 002 stance identification was mostly cast as a classi-003 fication problem. However, stance decisions in-004 volve complex judgments, which can be nowa-005 days
Maxwell Weinzierl, S. Harabagiu
semanticscholar   +1 more source

Autonomous Data Selection with Zero-shot Generative Classifiers for Mathematical Texts

Annual Meeting of the Association for Computational Linguistics
We present Autonomous Data Selection (AutoDS), a method that leverages base language models themselves as zero-shot"generative classifiers"to automatically curate high-quality mathematical texts.
Yifan Zhang, Yifan Zhang, Yifan Luo, Yang Yuan, A. C. Yao   +4 more
semanticscholar   +1 more source

Linguistic Analysis of Laser Speckle Contrast Images Recorded at Rest and During Biological Zero: Comparison With Laser Doppler Flowmetry Data

IEEE Transactions on Medical Imaging, 2013
Laser speckle contrast imaging (LSCI) is a newly commercialized imaging modality to monitor microvascular blood flow. Contrary to the well-known laser Doppler flowmetry (LDF), LSCI has the advantage of giving a full-field image of surface blood flow using simple instrumentation.
Anne, Humeau-Heurtier, Pierre, Abraham, Guillaume, Mahe   +2 more
openaire   +2 more sources

MobileSpeech: A Fast and High-Fidelity Framework for Mobile Zero-Shot Text-to-Speech

Annual Meeting of the Association for Computational Linguistics
Zero-shot text-to-speech (TTS) has gained significant attention due to its powerful voice cloning capabilities, requiring only a few seconds of unseen speaker voice prompts. However, all previous work has been developed for cloud-based systems.
Shengpeng Ji, Ziyue Jiang, Hanting Wang, Jialong Zuo, Zhou Zhao   +4 more
semanticscholar   +1 more source