On the Strength of Sequence Labeling and Generative Models for Aspect Sentiment Triplet Extraction.
Shen ZhouTieyun QianPublished in: ACL (Findings) (2023)
Keyphrases
- sequence labeling
- generative model
- conditional random fields
- discriminative models
- probabilistic model
- semi supervised learning
- discriminative learning
- information extraction
- sentiment analysis
- structured prediction
- named entity recognition
- dependency parsing
- maximum entropy
- markov models
- graphical models
- em algorithm
- machine learning
- learning tasks
- semi supervised
- prior knowledge
- image statistics
- hidden variables
- superpixels
- markov random field
- object categories
- unlabeled data
- image segmentation
- text classification
- pairwise
- learning algorithm