Filters








28 Hits in 7.0 sec

Camouflaged Chinese Spam Content Detection with Semi-supervised Generative Active Learning

Zhuoren Jiang, Zhe Gao, Yu Duan, Yangyang Kang, Changlong Sun, Qiong Zhang, Xiaozhong Liu
2020 Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics   unpublished
We propose a Semi-supervIsed GeNerative Active Learning (SIGNAL) model to address the imbalance, efficiency, and text camouflage problems of Chinese text spam detection task.  ...  To the best of our knowledge, this is the first work to integrate active learning and semisupervised generative learning for text spam detection.  ...  Conclusion In this paper, we propose a SIGNAL model for Chinese text spam detection. SIGNAL integrates active learning and semi-supervised generative learning into a unified framework.  ... 
doi:10.18653/v1/2020.acl-main.279 fatcat:vd5q67cgmnddrbqg3hk2gofk34

Social Fraud Detection Review: Methods, Challenges and Analysis [article]

Saeedreza Shehnepoor, Roberto Togneri, Wei Liu, Mohammed Bennamoun
2021 arXiv   pre-print
With this framework, a comprehensive overview of approaches is presented including supervised, semi-supervised, and unsupervised learning.  ...  The supervised approaches for fraud detection are introduced and categorized into two sub-categories; classical, and deep learning.  ...  Social Fraud in Comparison with Other Types of Spam Compared with other spam contents (e.g., email spam, insults, threats, malicious links, and fake news), fraud review detection is more challenging.  ... 
arXiv:2111.05645v1 fatcat:qp3zuv74lbaq3hw2ajxm6lfkim

Fraud Detection in Online Product Review Systems via Heterogeneous Graph Transformer

Songkai Tang, Luhua Jin, Fan Cheng
2021 IEEE Access  
This observation demonstrates the advan- tage of semi-supervised learning, where a small fraction of supervised signals is enough to optimize model parameters and generate informative node representation  ...  The whole model is trained in a semi-supervised manner. V.  ... 
doi:10.1109/access.2021.3084924 fatcat:wzzwnmdptnfm5hvarripls7heu

Fake Reviewer Group Detection in Online Review Systems [article]

Chen Cao, Shihao Li, Shuo Yu, Zhikui Chen
2021 arXiv   pre-print
Previous methods tackle this problem by detecting malicious individuals, ignoring the fact that the spam activities are often formed in groups, where individuals work collectively to write fake reviews  ...  First, cohensive groups are detected with modularity-based graph convolutional networks.  ...  Welling, “Semi-supervised classification with graph [39] A. Grover and J.  ... 
arXiv:2112.06403v1 fatcat:vf6ku3uezva23fuoi6kohnooay

Contextual Multi-View Query Learning for Short Text Classification in User-Generated Data [article]

Payam Karisani, Negin Karisani, Li Xiong
2021 arXiv   pre-print
We propose a novel multi-view active learning model, called Context-aware Co-testing with Bagging (COCOBA), to address these issues in the classification tasks tailored for a query word--e.g., detecting  ...  Mining user-generated content--e.g., for the early detection of outbreaks or for extracting personal observations--often suffers from the lack of enough training data, short document length, and informal  ...  Overcoming practical issues of deep active learn- 2020. Camouflaged Chinese spam content detection ing and its applications on named entity recognition.  ... 
arXiv:2112.02611v1 fatcat:t4c63auyqndwrpvx6xs3afeqoq

A review of machine learning approaches to Spam filtering

Thiago S. Guzella, Walmir M. Caminhas
2009 Expert systems with applications  
In this paper, we present a comprehensive review of recent developments in the application of machine learning algorithms to Spam filtering, focusing on both textual-and image-based approaches.  ...  Instead of considering Spam filtering as a standard classification problem, we highlight the importance of considering specific characteristics of the problem, especially concept drift, in designing new  ...  In the semi-supervised scenario, two classifiers were simultaneously trained, with the features being split for each one.  ... 
doi:10.1016/j.eswa.2009.02.037 fatcat:gf5z34w6arcdzh2w36tgefqppa

A Comprehensive Survey on Graph Anomaly Detection with Deep Learning [article]

Xiaoxiao Ma, Jia Wu, Shan Xue, Jian Yang, Chuan Zhou, Quan Z. Sheng, Hui Xiong, Leman Akoglu
2021 arXiv   pre-print
For the advent of deep learning, graph anomaly detection with deep learning has received a growing attention recently.  ...  , and social spam.  ...  When label information is available/partiallyavailable, supervised/semi-supervised detection models could be effectively trained.  ... 
arXiv:2106.07178v4 fatcat:efargsqnxndqbfqat2q5iz54u4

Online Social Deception and Its Countermeasures for Trustworthy Cyberspace: A Survey [article]

Zhen Guo, Jin-Hee Cho, Ing-Ray Chen, Srijan Sengupta, Michin Hong, Tanushree Mitra
2020 arXiv   pre-print
Cyber attackers have exploited the sophisticated features of SNSs to carry out harmful OSD activities, such as financial fraud, privacy threat, or sexual/labor exploitation.  ...  of social deception; (ii) types of OSD attacks and their unique characteristics compared to other social network attacks and cybercrimes; (iii) comprehensive defense mechanisms embracing prevention, detection  ...  spam to non-spam ratios and continuous sampling method, ground truth from commercial tool [21] Naïve Bayes, logistic regression, RF and semi- supervised spam detection Hashtag, content, user  ... 
arXiv:2004.07678v1 fatcat:k4a6siywefb6lhkmyn67lmoqwe

Online Social Deception and Its Countermeasures: A Survey

Zhen Guo, Jin-Hee Cho, Ing-Ray Chen, Srijan Sengupta, Michin Hong, Tanushree Mitra
2020 IEEE Access  
Based on this survey, we provide insights into the effectiveness of countermeasures and the lessons learned from the existing literature.  ...  Cyber attackers have exploited the sophisticated features of SNSs to carry out harmful OSD activities, such as financial fraud, privacy threat, or sexual/labor exploitation.  ...  [91] detected user profiles across multiple OSNs in a supervised learning classifier.  ... 
doi:10.1109/access.2020.3047337 fatcat:xw2rr2sjnrdf3nk4vfuowrkizy

Financial Cybercrime: A Comprehensive Survey of Deep Learning Approaches to Tackle the Evolving Financial Crime Landscape

Jack Nicholls, Aditya Kuppa, Nhien-An Le-Khac
2021 IEEE Access  
Welling, “Semi-supervised classification with graph //dx.doi.org/10.1016/j.jnca.2015.11.016 convolutional networks,” 5th International Conference on Learning Rep  ...  active learning,” 2020.  ... 
doi:10.1109/access.2021.3134076 fatcat:lm2upcaoabbnbie6r4sfzhjh4y

Machine Bias: Artificial Intelligence and Discrimination

Can Yavuz
2019 Social Science Research Network  
For instance, artificial intelligence is used to detect spam emails and to do so, it needs to learn the difference between spam and non-spam emails.  ...  Spam email fighting aims to detect spam emails and automatically move them into junk folder.  ... 
doi:10.2139/ssrn.3439702 fatcat:54zv7vvppne7dn6gcfh3phxt5e

Real-Time Detection Method for Surface Defects of Stamping Parts Based on Template Matching*

Bin Li, Yun Wu, Fengxia Guo, Jun Qi
2019 IOP Conference Series: Earth and Environment  
Studies, focusing on network security, have experienced four main stages: idealized design for ensuring security, auxiliary examination and passive defense, active analysis and strategy formulation, and  ...  He has been engaged in software development for 4 years in active network, and later served as general manager in medium-sized software enterprises. He has rich theoretical and practical experience.  ...  This method is generally associated with the alarm mechanism of IDS and is widely used in intrusion detection.  ... 
doi:10.1088/1755-1315/252/2/022076 fatcat:rqbvta44rrcivoyjzy5pebwdwe

Analysis framework of network security situational awareness and comparison of implementation methods

Yan Li, Guang-qiu Huang, Chun-zi Wang, Ying-chao Li
2019 EURASIP Journal on Wireless Communications and Networking  
Studies, focusing on network security, have experienced four main stages: idealized design for ensuring security, auxiliary examination and passive defense, active analysis and strategy formulation, and  ...  He has been engaged in software development for 4 years in active network, and later served as general manager in medium-sized software enterprises. He has rich theoretical and practical experience.  ...  This method is generally associated with the alarm mechanism of IDS and is widely used in intrusion detection.  ... 
doi:10.1186/s13638-019-1506-1 fatcat:lzamend3krbfhmntenrhmywx2u

D1.1 - State of the Art Analysis

Danilo Ardagna
2021 Zenodo  
The deliverable starts with an overview of AI applications and edge computing market trends.  ...  Then, the deliverable provides a background on AI applications design, also considering some advanced design trends (e.g., Network Architecture Search, Federated Learning, Deep Neural Networks partitioning  ...  A possible solution comes from a learning paradigm which sits between the unsupervised and supervised learning ones, which can be defined as semi-supervised learning.  ... 
doi:10.5281/zenodo.6372377 fatcat:f6ldfuwivbcltew4smiiwphfty

SIENNA D4.4: Ethical Analysis of AI and Robotics Technologies

Philip Jansen, Philip Brey, Alice Fox, Jonne Maas, Bradley Hillas, Nils Wagner, Patrick Smith, Isaac Oluoch, Laura Lamers, Hero Van Gein, Anaïs Resseguier, Rowena Rodrigues (+2 others)
2020 Zenodo  
The experience focuses on the algorithm's learning process and can be divided into supervised and unsupervised learning. 78 Supervised learning is the most common form.  ...  humans. 328, 329 In the supervised and semi-autonomous systems, the need to hand off control from robot to human at various points of operation present challenges with safety implications. 330 Some of  ... 
doi:10.5281/zenodo.4068082 fatcat:xiucqv6opng6rbit25lyfemzm4
« Previous Showing results 1 — 15 out of 28 results