ACL RD-TEC 1.0 Summarization of H01-1029
Paper Title:
FINE-GRAINED HIDDEN MARKOV MODELING FOR BROADCAST-NEWS STORY SEGMENTATION
FINE-GRAINED HIDDEN MARKOV MODELING FOR BROADCAST-NEWS STORY SEGMENTATION
Authors: W. Greiff and A. Morgan and R. Fish and M. Richards and A. Kundu
Primarily assigned technology terms:
- algorithm
- bayes classifier
- binning method
- categorization
- classifier
- density estimation
- dynamic programming
- hidden markov
- hidden markov model
- hidden markov modeling
- identification
- indexing
- iterbi
- kernel
- kernel density estimation
- kernel estimation
- kernels
- local context analysis
- markov model
- markov modeling
- maximum likelihood
- model development
- modeling
- naive bayes
- naive bayes classifier
- neural network
- news story segmentation
- non-parametric estimation
- non-parametric kernel
- non-parametric kernel estimation
- non-parametric kernel estimation techniques
- parameter estimation
- predictor
- recognizer
- regression
- segmentation
- segmentation algorithm
- speech recognizer
- splitting
- statistical analysis
- story segmentation
- tuning
- visualization
- viterbi
- viterbi algorithm
Other assigned terms:
- approach
- broadcast news
- buffer
- case
- clusters
- coherence
- conditional probabilities
- conditional probability
- conditional probability distribution
- distribution
- duration
- estimation
- evaluation metric
- fact
- feature
- feature value
- generation
- generative model
- histogram
- hmm model
- index
- interpretation
- joint distribution
- joint probability
- joint probability distribution
- language model
- likelihood
- local context
- measure
- method
- nist
- prior probability
- probabilities
- probability
- probability density
- probability distribution
- probability estimates
- segments
- sentence
- sentences
- signal
- speaker change
- statistics
- stem
- system performance
- technology
- test set
- tokens
- topology
- trained model
- training
- training data
- transcript
- transition probabilities
- trigram
- trigram model
- unigram
- word
- word-occurrence
- words
