Averaged Logits: An Weakly-Supervised Approach to Use Ratings to Train Sentence-Level Sentiment Classifiers.
Xiaoyu TangWei OuVan-Nam HuynhPublished in: IUKM (2019)
Keyphrases
- sentence level
- weakly supervised
- sentiment analysis
- sentiment classification
- topic models
- novelty detection
- opinion mining
- object detectors
- object class
- relation extraction
- semi supervised
- training data
- multi document summarization
- relevance model
- named entities
- text classification
- support vector
- decision trees
- training examples
- feature set
- co occurrence
- natural language processing
- object detection
- object recognition
- feature selection
- artificial intelligence
- image classification
- named entity recognition
- information extraction
- probabilistic model
- training set