Camouflaged Chinese Spam Content Detection with Semi-supervised Generative Active Learning.
Zhuoren JiangZhe GaoYu DuanYangyang KangChanglong SunQiong ZhangXiaozhong LiuPublished in: ACL (2020)
Keyphrases
- semi supervised
- active learning
- spam detection
- generative model
- semi supervised learning
- unsupervised learning
- labeled data
- unlabeled data
- pool based active learning
- supervised learning
- labelled training data
- pairwise
- multi view
- detection method
- false positives
- detection algorithm
- object detection
- semi supervised classification
- spam filtering
- labeled examples
- user generated content
- batch mode
- multimedia
- anti spam
- learning strategies
- co training
- training set
- social bookmarking systems
- learning algorithm
- semi supervised clustering
- metadata
- constrained clustering
- machine learning
- pairwise constraints
- named entity recognition
- cost sensitive
- data sets
- learning process
- information extraction