Open AccessResearch on sentiment analysis of short text based on Attention
Author(s) -
Haotian Li,
Hui Cao,
Zhang Xiaxia
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1601/5/052008
Subject(s) - computer science , artificial intelligence , artificial neural network , sentiment analysis , task (project management) , natural language processing , function (biology) , recurrent neural network , sequence (biology) , mechanism (biology) , machine learning , genetics , management , evolutionary biology , economics , biology , philosophy , epistemology
In order to classify the short essays, a circulating neural network is used to classify the texts by natural language processing technology. Recurrent neural network has some advantages in processing text sequence with memory function. In this paper, BiLSTM algorithm is used to extract text data features, and self-attention mechanism is introduced to improve the fitting ability of the model. The accuracy of this model algorithm on text emotion classification task is 78.1%, which is improved compared with LSTM and CNN neural network model.