Changchun Yang, Cong Cai*, Tongguang Ni
College of Computer and Artificial Science, Changzhou University, Changzhou, Jiangsu, China.
*Corresponding author: Cong Cai
Abstract
At present, the research on microblog short text has reached saturation, while the middle and long text sentences are complex and have many emotional words, which makes it difficult to classify the whole sentence. In order to solve the problem that the emotional feature extraction is not sufficient in the current micro-blog long text sentiment analysis task, which leads to the inability to extract the text sentiment semantics comprehensively, a multi-feature fusion sentiment analysis method (MBEA) combining capsule network, multi-layer bidirectional long short-term memory network and residual network is proposed. This method uses the Word2vec model to generate word vectors, and then inputs the word vectors into the residual network, capsule network and multi-layer bidirectional long-term and short-term memory network to obtain their vector feature representations respectively. Finally, the fully connected layer is input, and the emotion discrimination is performed by the softmax activation function. Experiments on COVID Dataset and Financial Dataset verify the accuracy and effectiveness of the model compared with other baseline models of sentiment analysis.
References
[1] Yan Zhongpei, Lu Wenxing, Shuang Juan, Wang Bingyou. A method for constructing an emotion dictionary for online travel-oriented reviews [J]. Computer Application Research, 2019, 36(06):1660-1664.
[2] Zhu Qinglin, Liang Bin, Xu Ruifeng, Liu Yuhan, Chen Yi, Mao Ruibin. Fine-grained sentiment analysis combining sentiment lexicon and attention mechanism in financial domain [J]. Chinese Journal of Information, 2022, 36(08):109-117.
[3] Kim Y. Convolutional neural networks for sentence classification [C]//Proceedings of The 2014 Conference on Empirical Method Natural Language Processing (EMNLP). Strouds burg, USA: Association for Computational Linguistics, 2014:1746-1751.
[4] Li Ping, Dai Yueming, Wu Dinghui. Application of dual-channel convolutional neural network in text sentiment analysis [J]. Computer application, 2018, 38(6): 1542-1546.
[5] Chen P, Xu B, Yang M, et al. Clause sentiment identification based on convolutional neural network with context embedding [C]//12th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD), IEEE, 2016:1532-1538.
[6] Wang J, Yu L C, Lai K R, et al. Dimensional sentiment analysis using a regional CNN-LSTM model[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), 2016: 225-230.
[7] Yang Yujuan, Yuan Huanhuan, Wang Yongli. A sentiment analysis method for review text [J]. Journal of Nanjing University of Science and Technology, 2019, 43(03): 280-285 + 291.
[8] Zhao W, Ye J, Yang M, et al. Investigating capsule networks with dynamic routing for text classification [J]. ArXiv Preprint ArXiv: 1804. 00538, 2018. Version 1.
[9] Zhang N, Deng S, Sun Z, et al. Attention-based capsule networks with dynamic routing for relation extraction [C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2018:986-992.
How to cite this paper
Sentiment Analysis of Medium and Long Text Based on Feature Fusion Model
How to cite this paper: Changchun Yang, Cong Cai, Tongguang Ni. (2023) Sentiment Analysis of Medium and Long Text Based on Feature Fusion Model. Advances in Computer and Communication, 4(2), 74-79.
DOI: http://dx.doi.org/10.26855/acc.2023.04.002