Abstract
For English toxic comment classification, this paper presents the model that combines Bi-GRU and CNN optimized by global average pooling (BG-GCNN) based on the bidirectional gated recurrent unit (Bi-GRU) and global pooling optimized convolution neural network (CNN). The model treats each type of toxic comment as a binary classification. First, Bi-GRU is used to extract the time-series features of the comment and then the dimensionality is reduced through global pooling optimized convolution neural network. Finally, the classification result is output by Sigmoid function. Comparative experiments show the BG-GCNN model has a better classification effect than Text-CNN, LSTM, Bi-GRU, and other models. The Macro-F1 value of the toxic comment dataset on the Kaggle competition platform is 0.62. The F1 values of the three toxic label classification results (toxic, obscene, and insult label) are 0.81, 0.84, and 0.74, respectively, which are the highest values in the comparative experiment.
- [1] 2015. Harassment Survey. Wikimedia Foundation, 2015. https://foundation.wikimedia.org/wiki/File:Harassment_Survey_2015_-_Results_Report.pdf.Google Scholar
- [2] . 2011. Modeling the detection of textual cyberbullying. In Fifth International AAAI Conference on Weblogs and Social Media.Google Scholar
- [3] . 2012. Learning from bullying traces in social media. In Proceedings of the 2012 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Association for Computational Linguistics, 656–666. Google Scholar
Digital Library
- [4] . 2017. Automated hate speech detection and the problem of offensive language. In Eleventh International AAAI Conference on Web and Social Media.Google Scholar
- [5] . 2018. Convolutional neural networks for toxic comment classification. In Proceedings of the 10th Hellenic Conference on Artificial Intelligence. 1–6. Google Scholar
Digital Library
- [6] . 2014. Convolutional neural networks for sentence classification. arXiv preprint arXiv:1408.5882.Google Scholar
- [7] . An Evaluation of Neural Network Models for Toxic Comment Classification.Google Scholar
- [8] . 2018. LSTM with attention for aggression detection. In Proceedings of the First Workshop on Trolling, Aggression and Cyberbullying (TRAC-2018). 52–57.Google Scholar
- [9] . 2018. TRAC-1 shared task on aggression identification: IIT (ISM)@ COLING’18. In Proceedings of the First Workshop on Trolling, Aggression and Cyberbullying (TRAC-2018). 58–65.Google Scholar
- [10] . 2019. Simple bidirectional LSTM solution for text classification. Proceedings of the Pol Eval 2019 Workshop, 2019: 111.Google Scholar
- [11] . 2018. Identifying aggression and toxicity in comments using capsule network. In Proceedings of the First Workshop on Trolling, Aggression and Cyberbullying (TRAC-2018). 98–105.Google Scholar
- [12] . 1998. Rethinking innateness: A connectionist perspective on development, Vol. 10. MIT Press.Google Scholar
- [13] . 1997. Long short-term memory. Neural Computation 9, 8 (1997), 1735–1780. Google Scholar
Digital Library
- [14] . 2016. Long short-term memory-networks for machine reading. arXiv preprint arXiv:1601.06733.Google Scholar
- [15] . 2020. Chinese speech recognition based on bi-directional circulatory neural network [J/OL]. Applied Acoustics, 2020(03):1–8 [2020-06-02]. http://kns.cnki.net/kcms/detail/11.2121.o4.20200506.1009.022.html.Google Scholar
- [16] . 2020. Discriminatory discriminations of converting bidirectional gated circulatory units and convolutional neural networks. Journal of Wuhan University (Science Edition) 66, 02 (2020), 111–116.Google Scholar
- [17] . 1998. Gradient-based learning applied to document recognition. Proceedings of the IEEE 86, 11 (1998), 2278–2324.Google Scholar
- [18] . 2013. Network in network. arXiv preprint arXiv:1312.4400.Google Scholar
- [19] 2016. Learning deep features for discriminative localization. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016: 2921–2929.Google Scholar
Cross Ref
- [20] . 2020. Analysis of complex cognitive task and pattern recognition using distributed patterns of EEG signals with cognitive functions. Neural Computing and Applications.
DOI: DOI: 10.1007/s00521-020-05439-9Google Scholar - [21] . 2020. Senti-eSystem: A sentiment-based eSystem -using hybridized fuzzy and deep neural network for measuring customer satisfaction. Software: Practice and Experience 51, 3 (2020), 571–594.
DOI: DOI: 10.1002/spe.2853Google ScholarCross Ref
- [22] . 2019. Segmentation methods for image classification using a convolutional neural network on AR-sandbox. IFIP Advances in Information and Communication Technology Artificial Intelligence Applications and Innovations. 391–398.
DOI: DOI: 10.1007/978-3-030-19823-7_33Google Scholar - [23] A framework for extractive text summarization based on deep learning modified neural network classifier. ACM Transactions on Asian and Low-Resource Language Information Processing 2020.
DOI: DOI: 10.1145/3392048 Google ScholarDigital Library
Index Terms
Toxic Comment Classification Based on Bidirectional Gated Recurrent Unit and Convolutional Neural Network
Recommendations
Indonesian Abstractive Text Summarization Using Bidirectional Gated Recurrent Unit
AbstractAbstractive text summarization is more challenging than the extractive one since it is performed by paraphrasing the entire contents of the text, which has a higher difficulty. But, it produces a more natural summary and higher inter-sentence ...
A Comparison Study of Convolutional Neural Network and Recurrent Neural Network on Image Classification
ICIT '22: Proceedings of the 2022 10th International Conference on Information Technology: IoT and Smart CityImage classification is a very important task in the field of computer vision, and it is widely used in daily life. In recent years, deep learning has developed rapidly in the field of image classification. Image classification methods based on deep ...
New GRU from convolutional neural network and gated recurrent unit
DATA '18: Proceedings of the First International Conference on Data Science, E-learning and Information SystemsThis paper describes our comparison between two deep learning systems. Initially we start with the first system which is used Convolutional Neural Network CNN which we will compare with the second system which is used Gated Recurrent Unit GRU. And ...






Comments