HMM-based Address Parsing with Massive Synthetic Training Data Generation.
Xiang LiHakan KardesXin WangAng SunPublished in: LocWeb (2014)
Keyphrases
- data generation
- hidden markov models
- active learning
- training process
- data streams
- multi dimensional
- low cost
- training examples
- unsupervised learning
- natural language
- data sets
- classification accuracy
- training set
- high throughput
- supervised learning
- model selection
- support vector
- streaming data
- context free grammars
- three dimensional