Classification of Government Consulting Departments Based on Deep Learning
DOI:
https://doi.org/10.54097/jceim.v10i2.6390Keywords:
Text classification, Deep Learning, LSTM, GRU, Bidirectional LSTM, Bidirectional GRU, Message from the massesAbstract
To further facilitate the resolution of problems encountered by the public in handling affairs, the government has opened many government service platforms. As a result of the large number of government departments, messages on platforms cannot be accurately and timely divided into various departments. Based on messages on the national government service platform, this paper establishes a deep learning model to classify messages. Sixteen government departments were involved in the analysis and the study found that the operational boundaries between some departments were not clear. In this paper, we analyzed the departments with more misclassifications and proposed targeted improvement solutions. Since there is a business transfer between the Bureau of Administrative Approval Service and the Bureau of Housing and Urban-Rural Development, we will not consider it in the subsequent analysis. In this paper, LSTM, bidirectional LSTM, GRU and bidirectional GRU were used to classify messages on the platform in 14 departments, with an accuracy of more than 80%. Among them, the accuracy of bidirectional GRU was the highest, and the speed of GRU and bidirectional GRU was fast, which realized the rapid and accurate classification of mass problems to the relevant government departments.
References
Minaee S, Kalchbrenner N, Cambria E, et al. Deep Learning Based Text Classification: A Comprehensive Review[J]. 2020.
Zhou C, Sun C, Liu Z, et al. A C-LSTM Neural Network for Text Classification[J]. Computer Science, 1(4):39-44, 2015.
A. H M, S. C. Computer network intrusion detection using sequential LSTM Neural Networks autoencoders[C]// The Proceedings of 26th Signal Processing and Communications Applications Conference (SIU). IEEE 1-4, 2018.
Cho K, Van Merriënboer B, Gulcehre C, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[J]. arXiv preprint arXiv:1406. 1078, 2014.
K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778, 2016.
Peng Z, Wei S, Tian J, et al. Attention-Based Bidirectional Long Short-Term Memory Networks for Relation Classification[C]// Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers). 2016.
Downloads
Published
Issue
Section
How to Cite
