Semantic and Generative Models for Lossy Text Compression.
Ian H. WittenTimothy C. BellAlistair MoffatCraig G. Nevill-ManningTony C. SmithHarold W. ThimblebyPublished in: Comput. J. (1994)
Keyphrases
- generative model
- text compression
- compression scheme
- probabilistic model
- mixture model
- variable length
- discriminative learning
- em algorithm
- discriminative models
- conditional random fields
- data compression
- semi supervised
- hierarchical hidden markov models
- natural language
- maximum entropy principle
- generative and discriminative models
- prior knowledge
- computer vision
- high quality
- object categories
- natural language processing
- hidden markov models
- topic models
- representational power
- probability distribution