Irony is a method of expressing sentiment implicitly.Differences between the words and the emotions of iron-ic sentences are abundant,causing difficulty in the sentiment classification of ironic sentences.To solve this problem,an ironic sentence recognition model integrating ironic language features(ISR)is proposed to improve the recognition ac-curacy of the ironic sentence by adding ironic language features.Initially,the Chi-square test algorithm is used to ana-lyze ironic language and obtain language features.Then,Word2Vec is used to train the language features to obtain the feature representation of the language features.At the same time,the attention mechanism and Bi-GRU(bidirectional gated recursive neural unit)model are used to obtain the feature representation of the sentence.Finally,the feature rep-resentations of language features and sentences are fused as the input of the sentiment classification layer to identify the ironic sentences.The model has been compared with CNN-AT,CNN-Adv,and EPSN models.Experiment results show that the proposed model has high recognition accuracy for the ironic sentence.
维普
万方数据