Social Bots Detection via Fusing BERT and Graph Convolutional Networks
The online social media ecosystem is becoming more and more confused because of more and more fake information and the social media of malicious users’ fake content; at the same time, unspeakable pain has been brought to mankind. Social robot detection uses supervised classification based on artific...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-12-01
|
| Series: | Symmetry |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-8994/14/1/30 |
| _version_ | 1797490147585949696 |
|---|---|
| author | Qinglang Guo Haiyong Xie Yangyang Li Wen Ma Chao Zhang |
| author_facet | Qinglang Guo Haiyong Xie Yangyang Li Wen Ma Chao Zhang |
| author_sort | Qinglang Guo |
| collection | DOAJ |
| description | The online social media ecosystem is becoming more and more confused because of more and more fake information and the social media of malicious users’ fake content; at the same time, unspeakable pain has been brought to mankind. Social robot detection uses supervised classification based on artificial feature extraction. However, user privacy is also involved in using these methods, and the hidden feature information is also ignored, such as semi-supervised algorithms with low utilization rates and graph features. In this work, we symmetrically combine BERT and GCN (Graph Convolutional Network, GCN) and propose a novel model that combines large scale pretraining and transductive learning for social robot detection, BGSRD. BGSRD constructs a heterogeneous graph over the dataset and represents Twitter as nodes using BERT representations. Corpus learning via text graph convolution network is a single text graph, which is mainly built for corpus-based on word co-occurrence and document word relationship. BERT and GCN modules can be jointly trained in BGSRD to achieve the best of merit, training data and unlabeled test data can spread label influence through graph convolution and can be carried out in the large-scale pre-training of massive raw data and the transduction learning of joint learning representation. The experiment shows that a better performance can also be achieved by BGSRD on a wide range of social robot detection datasets. |
| first_indexed | 2024-03-10T00:26:51Z |
| format | Article |
| id | doaj.art-9a975943013f4e13a1f360c2b4cdfb2c |
| institution | Directory Open Access Journal |
| issn | 2073-8994 |
| language | English |
| last_indexed | 2024-03-10T00:26:51Z |
| publishDate | 2021-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Symmetry |
| spelling | doaj.art-9a975943013f4e13a1f360c2b4cdfb2c2023-11-23T15:32:18ZengMDPI AGSymmetry2073-89942021-12-011413010.3390/sym14010030Social Bots Detection via Fusing BERT and Graph Convolutional NetworksQinglang Guo0Haiyong Xie1Yangyang Li2Wen Ma3Chao Zhang4School of Cyber Science and Technology, University of Science and Technology of China, Hefei 230027, ChinaSchool of Cyber Science and Technology, University of Science and Technology of China, Hefei 230027, ChinaNational Engineering Research Center for Public Safety Risk Perception and Control by Big Data (RPP), China Academic of Electronics and Information Technology, Beijing 100041, ChinaSchool of Software, Xinjiang University, Urumqi 830049, ChinaNational Engineering Research Center for Public Safety Risk Perception and Control by Big Data (RPP), China Academic of Electronics and Information Technology, Beijing 100041, ChinaThe online social media ecosystem is becoming more and more confused because of more and more fake information and the social media of malicious users’ fake content; at the same time, unspeakable pain has been brought to mankind. Social robot detection uses supervised classification based on artificial feature extraction. However, user privacy is also involved in using these methods, and the hidden feature information is also ignored, such as semi-supervised algorithms with low utilization rates and graph features. In this work, we symmetrically combine BERT and GCN (Graph Convolutional Network, GCN) and propose a novel model that combines large scale pretraining and transductive learning for social robot detection, BGSRD. BGSRD constructs a heterogeneous graph over the dataset and represents Twitter as nodes using BERT representations. Corpus learning via text graph convolution network is a single text graph, which is mainly built for corpus-based on word co-occurrence and document word relationship. BERT and GCN modules can be jointly trained in BGSRD to achieve the best of merit, training data and unlabeled test data can spread label influence through graph convolution and can be carried out in the large-scale pre-training of massive raw data and the transduction learning of joint learning representation. The experiment shows that a better performance can also be achieved by BGSRD on a wide range of social robot detection datasets.https://www.mdpi.com/2073-8994/14/1/30social bots detectGNNGCNpre-trainingBERT |
| spellingShingle | Qinglang Guo Haiyong Xie Yangyang Li Wen Ma Chao Zhang Social Bots Detection via Fusing BERT and Graph Convolutional Networks Symmetry social bots detect GNN GCN pre-training BERT |
| title | Social Bots Detection via Fusing BERT and Graph Convolutional Networks |
| title_full | Social Bots Detection via Fusing BERT and Graph Convolutional Networks |
| title_fullStr | Social Bots Detection via Fusing BERT and Graph Convolutional Networks |
| title_full_unstemmed | Social Bots Detection via Fusing BERT and Graph Convolutional Networks |
| title_short | Social Bots Detection via Fusing BERT and Graph Convolutional Networks |
| title_sort | social bots detection via fusing bert and graph convolutional networks |
| topic | social bots detect GNN GCN pre-training BERT |
| url | https://www.mdpi.com/2073-8994/14/1/30 |
| work_keys_str_mv | AT qinglangguo socialbotsdetectionviafusingbertandgraphconvolutionalnetworks AT haiyongxie socialbotsdetectionviafusingbertandgraphconvolutionalnetworks AT yangyangli socialbotsdetectionviafusingbertandgraphconvolutionalnetworks AT wenma socialbotsdetectionviafusingbertandgraphconvolutionalnetworks AT chaozhang socialbotsdetectionviafusingbertandgraphconvolutionalnetworks |