SUDMAD: Sequential and unsupervised decomposition of a multi-author document based on a hidden markov model.
Khaled AldebeiXiangjian HeWenjing JiaWei-Chang YehPublished in: J. Assoc. Inf. Sci. Technol. (2018)
Keyphrases
- hidden markov models
- sequential data
- speech recognition
- markov models
- information retrieval
- viterbi algorithm
- document analysis
- conditional random fields
- markov model
- baum welch
- unsupervised learning
- document clustering
- gesture recognition
- handwritten word recognition
- document images
- web documents
- supervised learning
- hierarchical hidden markov model
- hidden state
- multi stream
- vector space model
- dynamic probabilistic networks
- fixed length
- probabilistic model
- image sequences