ID	原文	译文
57478	348% ，召回率和 F1 值分别达到了 47.	348% . In addition，the recall and the F1 value of the HFF reach 47.
57479	8% 和 49.	8% and 49.
57480	4% ，优于现有可用于商品文本标题分类的先进短文本分类方法.	4% ，respectively，which is supe- rior to the existing advanced short text classification method that can be used for the commodity text titles classification.
57481	多导睡眠仪中的多种呼吸信号使睡眠呼吸暂停检测过程变得复杂，影响患者睡眠，对此，提出一种利用随机森林方法进行自动睡眠呼吸暂停检测的方法.	To solve the problem that various respiratory signals in polysomnography make the detection of sleep apnea complicated and affect patients＇sleep，a method of automatic sleep apnea detection using ran- dom forest is proposed.
57482	腹部呼吸信号经过希尔伯特－黄变换后，呼吸暂停与正常睡眠时的能量和边际谱分布显著不同，通过提取相关频域特征，结合时域特征，利用机器学习中的随机森林方法进行呼吸暂停的检测，有效地降低了检测复杂性，提高了检测准确性.	The energy and marginal spectrum distribution of sleep apnea is significantly dif- ferent from that of normal sleep after Hilbert-Huang transform. By extracting the relevant frequency do- main features，combining with the time domain features，the random forest method in machine learning method is used to detect sleep apnea，which effectively reduces the detection complexity and improve the accuracy.
57483	实验结果表明，此方法在检测的便捷性和准确性上优于已有方法，更适用于家庭环境，具有广泛的应用前景.	Experiments show that this method is more convenient and accurate than the existing method， more suitable for home environment，and has a wide range of application prospects.
57484	针对社交网络用户态度分析任务中用户之间原有社交关系方向可能阻碍态度信息流动以及标签扩散的问题，提出了一种应用于半监督图卷积网络的社交关系方向门控算法.	Facing the problem in social user attitude analysis that the original social relationship direction between users in social networks hinders the flow of attitude information and label propagation，a social relationship direction gating algorithm for graph convolutional networks is proposed.
57485	该算法首先在原有与逆向社交关系方向上分别进行图卷积运算，得到 2 种用户节点态度特征向量，然后利用门控机制对 2 种特征向量进行动态融合.	The algorithm first performs graph convolution on the origin and reverse social relationship directions to obtain two types of user node attitude feature vectors，and then leverages the gating mechanism to integrate the feature vectors dynamically.
57486	扩展了态度信息传播路径的同时，还能够捕捉用户影响力差异，以自动选择态度信息的流动方向.	While expanding the propagation of attitude information，the algorithm can also capture the differences of user influence to automatically select the flow of attitude information.
57487	在 2 个真实热点话题数据集上的实验结果表明，现有图卷积网络在加入该算法之后，其用户态度分析的准确率能够得到有效提升.	Experiments on two real hot topic datasets show that the accuracy of the existing graph convolutional networks can be ef- fectively improved after adding the proposed algorithm.