Semi-supervised Sentiment Analysis for Chinese Stock Texts in Scarce Labeled Data Scenario and Price Prediction.
Ji ZhaoyanYan HongfeiYing SipingChong ChenQi SuPublished in: CCIR (2020)
Keyphrases
- sentiment analysis
- labeled data
- semi supervised
- text classification
- unlabeled data
- semi supervised learning
- opinion mining
- text documents
- sentiment classification
- semi supervised classification
- labeled and unlabeled data
- active learning
- pairwise
- supervised learning
- supervised classifiers
- transfer learning
- domain adaptation
- labeled training data
- text categorization
- fully supervised
- training data
- co training
- labeled examples
- class labels
- semi supervised learning algorithms
- semi supervised learning methods
- text mining
- real world
- training examples
- sentence level
- feature selection
- pairwise constraints
- natural language processing
- label information
- learning algorithm
- data points
- metric learning
- semi supervised clustering
- unlabeled samples
- supervised methods
- data analysis
- cross lingual
- label propagation
- machine learning
- knn
- unsupervised learning
- data sets