为更好地解决传统模型特征捕捉能力不足,词向量语义表示不准确等问题,提出了结合MacBERT与多层次特征协同网络的音乐社交评论情感分析模型MacBERT-MFCN(MacBERT and Multi-level Feature Collaborative Network)。采用MacBERT模型提取...为更好地解决传统模型特征捕捉能力不足,词向量语义表示不准确等问题,提出了结合MacBERT与多层次特征协同网络的音乐社交评论情感分析模型MacBERT-MFCN(MacBERT and Multi-level Feature Collaborative Network)。采用MacBERT模型提取评论文本特征向量,解决静态词向量无法表示多义词的问题;多层次特征协同网络结合双向内置注意力简单循环单元(Bidirectional Built in Attention Simple Recurrent Unit,BiBASRU)和多层次卷积神经网络(Multilevel Convolutional Neural Network,MCNN)模块,全面捕捉局部和上下文语义特征;软注意力用来衡量分类特征贡献的大小,赋予关键特征更高权重。基于网易云评论文本数据集进行实验,结果表明,MacBERTMFCN模型F1值高达95.56%,能有效地提升文本情感分类准确率。展开更多
The rapid development of network technology and its evolution toward heterogeneous networks has increased the demand to support automatic monitoring and the management of heterogeneous wireless communication networks....The rapid development of network technology and its evolution toward heterogeneous networks has increased the demand to support automatic monitoring and the management of heterogeneous wireless communication networks.This paper presents a multilevel pattern mining architecture to support automatic network management by discovering interesting patterns from telecom network monitoring data.This architecture leverages and combines existing frequent itemset discovery over data streams,association rule deduction,frequent sequential pattern mining,and frequent temporal pattern mining techniques while also making use of distributed processing platforms to achieve high-volume throughput.展开更多
文摘为更好地解决传统模型特征捕捉能力不足,词向量语义表示不准确等问题,提出了结合MacBERT与多层次特征协同网络的音乐社交评论情感分析模型MacBERT-MFCN(MacBERT and Multi-level Feature Collaborative Network)。采用MacBERT模型提取评论文本特征向量,解决静态词向量无法表示多义词的问题;多层次特征协同网络结合双向内置注意力简单循环单元(Bidirectional Built in Attention Simple Recurrent Unit,BiBASRU)和多层次卷积神经网络(Multilevel Convolutional Neural Network,MCNN)模块,全面捕捉局部和上下文语义特征;软注意力用来衡量分类特征贡献的大小,赋予关键特征更高权重。基于网易云评论文本数据集进行实验,结果表明,MacBERTMFCN模型F1值高达95.56%,能有效地提升文本情感分类准确率。
基金funded by the Enterprise Ireland Innovation Partnership Programme with Ericsson under grant agreement IP/2011/0135[6]supported by the National Natural Science Foundation of China(No.61373131,61303039,61232016,61501247)+1 种基金the PAPDCICAEET funds
文摘The rapid development of network technology and its evolution toward heterogeneous networks has increased the demand to support automatic monitoring and the management of heterogeneous wireless communication networks.This paper presents a multilevel pattern mining architecture to support automatic network management by discovering interesting patterns from telecom network monitoring data.This architecture leverages and combines existing frequent itemset discovery over data streams,association rule deduction,frequent sequential pattern mining,and frequent temporal pattern mining techniques while also making use of distributed processing platforms to achieve high-volume throughput.
