评论是消费者对商品评价和反馈的一种文本形式。评论摘要是指对评论进行提取和压缩,形成能够概括评论信息的短文本。目前,评论摘要任务大多只关注评论的文本序列,忽略了评论中的方面、意见短语和情感极性等相关评价对象信息。因此,提出...评论是消费者对商品评价和反馈的一种文本形式。评论摘要是指对评论进行提取和压缩,形成能够概括评论信息的短文本。目前,评论摘要任务大多只关注评论的文本序列,忽略了评论中的方面、意见短语和情感极性等相关评价对象信息。因此,提出了一种基于T5模型(Text-to-Text Transfer Transformer),结合评价对象信息的评论摘要方法。该方法首先利用T5模型对评论摘要任务进行建模,通过注意力机制学习评论文本中的上下文信息,生成包含核心语义的摘要文本;然后提取摘要文本中的评价对象信息,并将其作为评论摘要任务的辅助信息;最后利用少样本数据对模型参数进行特异性调整,进一步改善摘要的效果,从而生成高质量的评论摘要。实验结果表明,在酒店评论数据集SPACE和产品评论数据集OPOSUM+上,该方法相较于基准模型在ROUGE评价指标上均有显著提升。展开更多
商品属性分类任务是指对一段商品的描述文字进行属性分析并进而对多个属性进行分类的过程,其有助于人们从多个角度了解商品,为市场营销、产品管理等提供帮助。当前大语言模型的使用也愈加广泛,但在商品属性分类问题上,通用大模型由于缺...商品属性分类任务是指对一段商品的描述文字进行属性分析并进而对多个属性进行分类的过程,其有助于人们从多个角度了解商品,为市场营销、产品管理等提供帮助。当前大语言模型的使用也愈加广泛,但在商品属性分类问题上,通用大模型由于缺乏领域知识和属性关联等信息,性能不够理想。为此,提出了一个基于双重预训练的商品属性分类方法,旨在通过使用特定的预训练方式提高大语言模型在商品属性分类任务中的性能。在T5模型的基础上,引入了领域内文本预训练和基于属性间关联性的预训练两种方法。在Clothing Fit Data数据集上的实验结果显示,使用了双重预训练的T5模型较未经过预训练的模型以及其他基准模型,在各个属性上的分类效果都取得了一定提升。实验结果证明了所提方法的有效性。展开更多
This study aims to address the deviation in downstream tasks caused by inaccurate recognition results when applying Automatic Speech Recognition(ASR)technology in the Air Traffic Control(ATC)field.This paper presents ...This study aims to address the deviation in downstream tasks caused by inaccurate recognition results when applying Automatic Speech Recognition(ASR)technology in the Air Traffic Control(ATC)field.This paper presents a novel cascaded model architecture,namely Conformer-CTC/Attention-T5(CCAT),to build a highly accurate and robust ATC speech recognition model.To tackle the challenges posed by noise and fast speech rate in ATC,the Conformer model is employed to extract robust and discriminative speech representations from raw waveforms.On the decoding side,the Attention mechanism is integrated to facilitate precise alignment between input features and output characters.The Text-To-Text Transfer Transformer(T5)language model is also introduced to handle particular pronunciations and code-mixing issues,providing more accurate and concise textual output for downstream tasks.To enhance the model’s robustness,transfer learning and data augmentation techniques are utilized in the training strategy.The model’s performance is optimized by performing hyperparameter tunings,such as adjusting the number of attention heads,encoder layers,and the weights of the loss function.The experimental results demonstrate the significant contributions of data augmentation,hyperparameter tuning,and error correction models to the overall model performance.On the Our ATC Corpus dataset,the proposed model achieves a Character Error Rate(CER)of 3.44%,representing a 3.64%improvement compared to the baseline model.Moreover,the effectiveness of the proposed model is validated on two publicly available datasets.On the AISHELL-1 dataset,the CCAT model achieves a CER of 3.42%,showcasing a 1.23%improvement over the baseline model.Similarly,on the LibriSpeech dataset,the CCAT model achieves a Word Error Rate(WER)of 5.27%,demonstrating a performance improvement of 7.67%compared to the baseline model.Additionally,this paper proposes an evaluation criterion for assessing the robustness of ATC speech recognition systems.In robustness evaluation experiments based on this criterion,the proposed model demonstrates a performance improvement of 22%compared to the baseline model.展开更多
By employing the improved T42L5 spectral model and utilizing the ECMWF data covering the period from 1 July to 7 July 1982,a numerical research on the formation of the Ural blocking system has been made.The results sh...By employing the improved T42L5 spectral model and utilizing the ECMWF data covering the period from 1 July to 7 July 1982,a numerical research on the formation of the Ural blocking system has been made.The results show that the model forecasts for the upstream U ral area turn out to be worse if the dynamic effect of the Qinghai-Xizang Plateau is not considered.The correlation coefficient between the model forecasts and observed 500 hPa geopotential height anomaly decreases by 9% for the 5-day mean,and their averaged root mean square (RMS) error increases 15 m.Due to the dynamic effect of the Plateau,the trough being on the northwest of the Plateau is barricaded and turns to be a transversal trough.Consequently southwest flow occurs along the northwest of the Plateau in front of the trough,while northeast flow prevails over the west of the trough,causing the formation of the blocking high over the Ural area.When the dynamic effect of the Plateau is not taken into consideration,the trough develops and moves southeastward and the Ural blocking high changes into a migratory high.All these result in the failure of the simulation.The dynamic effect of the Plateau helps to increase the negative vorticities over the Plateau and its north periphery as well as the Ural area,and also helps to increase the positive vorticities over the Black Sea and the Caspian Sea area.On the other hand,the thermodynamic effect mainly influences the Plateau and its downstream area and plays an less important role in the formation of the blocking high over the upstream Ural area.展开更多
文摘评论是消费者对商品评价和反馈的一种文本形式。评论摘要是指对评论进行提取和压缩,形成能够概括评论信息的短文本。目前,评论摘要任务大多只关注评论的文本序列,忽略了评论中的方面、意见短语和情感极性等相关评价对象信息。因此,提出了一种基于T5模型(Text-to-Text Transfer Transformer),结合评价对象信息的评论摘要方法。该方法首先利用T5模型对评论摘要任务进行建模,通过注意力机制学习评论文本中的上下文信息,生成包含核心语义的摘要文本;然后提取摘要文本中的评价对象信息,并将其作为评论摘要任务的辅助信息;最后利用少样本数据对模型参数进行特异性调整,进一步改善摘要的效果,从而生成高质量的评论摘要。实验结果表明,在酒店评论数据集SPACE和产品评论数据集OPOSUM+上,该方法相较于基准模型在ROUGE评价指标上均有显著提升。
文摘商品属性分类任务是指对一段商品的描述文字进行属性分析并进而对多个属性进行分类的过程,其有助于人们从多个角度了解商品,为市场营销、产品管理等提供帮助。当前大语言模型的使用也愈加广泛,但在商品属性分类问题上,通用大模型由于缺乏领域知识和属性关联等信息,性能不够理想。为此,提出了一个基于双重预训练的商品属性分类方法,旨在通过使用特定的预训练方式提高大语言模型在商品属性分类任务中的性能。在T5模型的基础上,引入了领域内文本预训练和基于属性间关联性的预训练两种方法。在Clothing Fit Data数据集上的实验结果显示,使用了双重预训练的T5模型较未经过预训练的模型以及其他基准模型,在各个属性上的分类效果都取得了一定提升。实验结果证明了所提方法的有效性。
基金This study was co-supported by the National Key R&D Program of China(No.2021YFF0603904)National Natural Science Foundation of China(U1733203)Safety Capacity Building Project of Civil Aviation Administration of China(TM2019-16-1/3).
文摘This study aims to address the deviation in downstream tasks caused by inaccurate recognition results when applying Automatic Speech Recognition(ASR)technology in the Air Traffic Control(ATC)field.This paper presents a novel cascaded model architecture,namely Conformer-CTC/Attention-T5(CCAT),to build a highly accurate and robust ATC speech recognition model.To tackle the challenges posed by noise and fast speech rate in ATC,the Conformer model is employed to extract robust and discriminative speech representations from raw waveforms.On the decoding side,the Attention mechanism is integrated to facilitate precise alignment between input features and output characters.The Text-To-Text Transfer Transformer(T5)language model is also introduced to handle particular pronunciations and code-mixing issues,providing more accurate and concise textual output for downstream tasks.To enhance the model’s robustness,transfer learning and data augmentation techniques are utilized in the training strategy.The model’s performance is optimized by performing hyperparameter tunings,such as adjusting the number of attention heads,encoder layers,and the weights of the loss function.The experimental results demonstrate the significant contributions of data augmentation,hyperparameter tuning,and error correction models to the overall model performance.On the Our ATC Corpus dataset,the proposed model achieves a Character Error Rate(CER)of 3.44%,representing a 3.64%improvement compared to the baseline model.Moreover,the effectiveness of the proposed model is validated on two publicly available datasets.On the AISHELL-1 dataset,the CCAT model achieves a CER of 3.42%,showcasing a 1.23%improvement over the baseline model.Similarly,on the LibriSpeech dataset,the CCAT model achieves a Word Error Rate(WER)of 5.27%,demonstrating a performance improvement of 7.67%compared to the baseline model.Additionally,this paper proposes an evaluation criterion for assessing the robustness of ATC speech recognition systems.In robustness evaluation experiments based on this criterion,the proposed model demonstrates a performance improvement of 22%compared to the baseline model.
文摘By employing the improved T42L5 spectral model and utilizing the ECMWF data covering the period from 1 July to 7 July 1982,a numerical research on the formation of the Ural blocking system has been made.The results show that the model forecasts for the upstream U ral area turn out to be worse if the dynamic effect of the Qinghai-Xizang Plateau is not considered.The correlation coefficient between the model forecasts and observed 500 hPa geopotential height anomaly decreases by 9% for the 5-day mean,and their averaged root mean square (RMS) error increases 15 m.Due to the dynamic effect of the Plateau,the trough being on the northwest of the Plateau is barricaded and turns to be a transversal trough.Consequently southwest flow occurs along the northwest of the Plateau in front of the trough,while northeast flow prevails over the west of the trough,causing the formation of the blocking high over the Ural area.When the dynamic effect of the Plateau is not taken into consideration,the trough develops and moves southeastward and the Ural blocking high changes into a migratory high.All these result in the failure of the simulation.The dynamic effect of the Plateau helps to increase the negative vorticities over the Plateau and its north periphery as well as the Ural area,and also helps to increase the positive vorticities over the Black Sea and the Caspian Sea area.On the other hand,the thermodynamic effect mainly influences the Plateau and its downstream area and plays an less important role in the formation of the blocking high over the upstream Ural area.
