From Conditional Random Field (CRF) to Rhetorical Structure Theory(RST): Incorporating Context Information in Sentiment Analysis.
Aggeliki VlachostergiouGeorge MarandianosStefanos D. KolliasPublished in: ESWC (Satellite Events) (2017)
Keyphrases
- conditional random fields
- sentiment analysis
- rhetorical structure theory
- crf model
- opinion mining
- hidden markov models
- text classification
- information extraction
- graphical models
- sentiment classification
- random fields
- probabilistic model
- generative model
- maximum entropy
- higher order
- natural language processing
- text mining
- sentence level
- product reviews
- markov random field
- markov networks
- pairwise
- semantic segmentation
- structured prediction
- contextual features
- parameter learning
- named entity recognition
- data mining
- segmentation method
- sentiment lexicon
- machine learning
- k means
- superpixels
- active learning
- bayesian networks
- public opinion
- online reviews
- ordinal scale