Supervised spoken document summarization based on structured support vector machine with utterance clusters as hidden variables.
Sz-Rung ShiangHung-yi LeeLin-Shan LeePublished in: INTERSPEECH (2013)
Keyphrases
- hidden variables
- document summarization
- support vector machine
- bayesian networks
- probabilistic model
- bayesian inference
- multi document summarization
- latent variables
- generative model
- em algorithm
- clustering algorithm
- feature selection
- supervised learning
- document set
- automatic summarization
- support vector
- cosine similarity
- unsupervised learning
- document clustering
- machine learning
- query specific
- learning algorithm
- kernel methods
- semantic features
- missing values
- syntactic analysis
- training data
- feature vectors
- knn
- data points
- posterior distribution
- incomplete data
- maximum likelihood
- image segmentation
- expectation maximization
- semantic information