ACL RD-TEC 1.0 Summarization of N06-2014
Paper Title:
AGREEMENT/DISAGREEMENT CLASSIFICATION: EXPLOITING UNLABELED DATA USING CONTRAST CLASSIFIERS
AGREEMENT/DISAGREEMENT CLASSIFICATION: EXPLOITING UNLABELED DATA USING CONTRAST CLASSIFIERS
Authors: Sangyun Hahn and Richard Ladner and Mari Ostendorf
Primarily assigned technology terms:
- bayesian approach
- bayesian network
- classi cation
- classi er
- classication
- clustering
- clustering technique
- co-training
- database
- decision tree
- iterative training
- language processing
- language understanding
- learning
- learning algorithms
- learning methods
- logistic regression
- meeting recording
- natural language understanding
- neural net
- outlier detection
- processing
- recognizer
- regression
- resampling
- self-training
- semi-supervised classication
- semi-supervised learning
- semi-supervised training
- support vector machine
- unsupervised clustering
Other assigned terms:
- adjacency pair
- approach
- bias
- bigram
- case
- chunk
- class distribution
- class probability
- classi cation accuracy
- conversational speech
- data set
- data sets
- distribution
- duration
- experimental results
- f measure
- feature
- feature set
- feature sets
- icsi meeting corpus
- labeling
- language model
- language models
- language understanding research
- lexical features
- likelihood
- measure
- meeting corpus
- meeting recording corpus
- method
- natural language
- pause
- perplexity
- posterior
- posterior probability
- probability
- process
- segment duration
- speaker change
- speech act
- support vector
- technique
- terms
- test set
- tokens
- training
- training data
- training set
- transcripts
- tree
- understanding
- word
- words
