基于BiLSTM和CNN的序贯三支情感分类模型研究

赵梦宇, 孙京博, 魏遵天, 辛现伟, 宋继华

Research on sequential three⁃way sentiment classification model based on BiLSTM and CNN

Mengyu Zhao, Jingbo Sun, Zuntian Wei, Xianwei Xin, Jihua Song

表2 三个数据集在不同模型下分类准确率对比

Table 2 Classification accuracy of three datasets with different models

模型 Hotel Computer 中文微博情感 BiLCNN⁃S3WD 0.9152±0.0022 0.9236±0.0050 0.8193±0.0049 BiLSTM 0.9123±0.0037 0.9082±0.0084 0.7977±0.0154 BiLSTM+CNN (kernel_size=2) 0.9120±0.0027 0.9194±0.0020 0.7583±0.0231 BiLSTM+CNN (kernel_size=3) 0.9129±0.0018 0.9200±0.0046 0.7999±0.0132 BiLSTM+CNN (kernel_size=4) 0.9139±0.0029 0.9206±0.0047 0.7939±0.0190 BiLSTM+CNN (kernel_size=5) 0.9130±0.0020 0.9192±0.0062 0.8095±0.0116 DLSTWSC 0.9106±0.0015 0.9230±0.0000 0.7855±0.0164 CFRT 0.8270±0.0023 0.8698±0.0018 0.8190±0.0012