相关参考文献
注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article
Biochemical Research Methods
Jiasheng Si et al.
Summary: Biomedical argument mining aims to identify and extract the argumentative structure in biomedical text, providing insights into medical decision making. Current approaches have achieved some success, but they ignore the sequential dependency between argument component classification and relation identification, as well as the valuable context for relation identification.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Article
Computer Science, Theory & Methods
Ziwei Ji et al.
Summary: Natural Language Generation (NLG) has made significant progress in recent years, thanks to deep learning technologies like Transformer-based language models. This advancement has resulted in more fluent and coherent NLG, leading to improvements in tasks such as summarization, dialogue generation, and data-to-text conversion. However, deep learning-based generation is prone to producing unintended text, which affects system performance and fails to meet user expectations in real-world scenarios. To address this issue, researchers have conducted studies on measuring and mitigating hallucinated texts. This survey provides an overview of the research progress and challenges in the hallucination problem in NLG, covering metrics, mitigation methods, and task-specific research progress.
ACM COMPUTING SURVEYS
(2023)
Article
Computer Science, Artificial Intelligence
Yanran Chen et al.
Summary: Recently proposed BERT-based evaluation metrics for text generation are susceptible to adversarial attacks due to being models of semantic similarity. In contrast, our evaluation metrics based on Natural Language Inference (NLI) show greater robustness to attacks. While our NLI metrics outperform existing summarization metrics, they still fall short of state-of-the-art machine translation metrics. However, combining existing metrics with our NLI metrics improves both adversarial robustness and quality metrics on standard benchmarks.
TRANSACTIONS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS
(2023)
Article
Computer Science, Artificial Intelligence
Philippe Laban et al.
Summary: This work revisits the use of natural language inference (NLI) models for inconsistency detection, and proposes a method called SummaC(Conv) that allows NLI models to be successfully applied to this task by segmenting documents into sentence units and aggregating scores between pairs of sentences. They also introduce a new benchmark called SummaC (Summary Consistency) consisting of six large inconsistency detection datasets. On this dataset, SummaC(Conv) achieves state-of-the-art results with a balanced accuracy of 74.4%, a 5% improvement compared to previous work.
TRANSACTIONS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS
(2022)
Article
Computer Science, Artificial Intelligence
Stefanos Angelidis et al.
Summary: The Quantized Transformer (QT) is an unsupervised system for extractive opinion summarization, inspired by Vector-Quantized Variational Autoencoders. QT utilizes the properties of the quantized space to drive popular opinion summaries and extract aspect-specific summaries, showing promise in the field of opinion summarization.
TRANSACTIONS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS
(2021)
Article
Computer Science, Artificial Intelligence
Giuseppe Carenini et al.
COMPUTATIONAL INTELLIGENCE
(2013)