Study on text classification model combining BERT and convolutional neural network
DOI:
https://doi.org/10.54097/7h5bs772Keywords:
Text classification,BERT-CNN,Dialect database.Abstract
Text classification has been widely used in many practical scenarios such as sentiment analysis and topic classification, and is an important task in the field of natural language processing. In recent years, research on pre-training model based on Bidirectional Encoder Representations from Transformers (BERT) has made remarkable progress. However, at present, most studies are based on open source databases. In this study, a text classification model combining BERT model and convolutional neural network (CNN) is proposed to complete the text emotion classification task through the self-built dialect text database. This paper demonstrates the effectiveness and superiority of this model in Chinese text classification tasks by comparing experiments and performance evaluation.
References
[1] Zhou Jun, He Li, Han Weihong, et al. Empirical research on propensity classification of Chinese comments based on Machine learning [J]. Information Network Security, 2013, (10):164-166.
[2] Lu Jiali. Multi-label classification method of open source threat intelligence text based on Bert-TextCNN [J]. Research on Information Security, 2018,10(08):760-768.
[3] Zou Liwei, He Zhi, Zhou Chengle. Typhoon disaster identification method based on Weibo and deep learning: A case study of the transit typhoon in Guangdong Province from 2010 to 2019 [J]. Tropical geography, 2024, 44 (6) : 1079-1089. The DOI: 10.13284 / j.carol carroll nki RDDL. 003882.
[4] Abas R A ,Elhenawy I ,Zidan M , et al.BERT-CNN: A Deep Learning Model for Detecting Emotions from Text[J]. Computers, Materials & Continua,2022,71(2):2943-2961.
[5] An Xin, Dai Zibiao, Li Yang, et al. End-to-end speech synthesis method based on BERT [J]. Computer Science, 202,49(04): 221-226. (in Chinese)
[6] Palace success. Sentiment analysis unsupervised adaptation research [D]. Nanjing university of science and technology, 2021. The DOI: 10.27241 /, dc nki. Gnjgu. 2021.001700.
[7] Xu Jianfei, Wu Yuecheng. Research on News text Classification based on BERT-BiLSTM-CNN model [J]. Software engineering, 2023, 26 (6) : 11-15. DOI: 10.19644 / j.carol carroll nki issn2096-1472.2023.006.003.
[8] Zhi Shiyao, Wu Zhenru, Chen Tao, et al. Research on Sentiment Analysis of microblog comments based on ALBERT-BiLSTM-Att [J]. The computer age, 2022, (02) : 19-22. DOI: 10.16644 / j.carol carroll nki cn33-1094 / tp. 2022.02.006.
[9] Yanling L ,Jiaye W ,Xudong L .BERT-CNN based evidence retrieval and aggregation for Chinese legal multi-choice question answering[J].Neural Computing and Applications, 2024,36(11):5909-5925.
[10] Bi X ,Zhang T .Pedagogical sentiment analysis based on the BERT-CNN-BiGRU-attention model in the context of intercultural communication barriers.[J].PeerJ. Computer science,2024,10e2166.
[11] Hamed K A ,Saleh F A ,Akhil K , et al.A novel multi-layer feature fusion-based BERT-CNN for sentence representation learning and classification[J].Robotic Intelligence and Automation,2023,43(6):704-715.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
