ACL RD-TEC 1.0 Summarization of W03-0402
Paper Title:
AN SVM-BASED VOTING ALGORITHM WITH APPLICATION TO PARSE RERANKING
AN SVM-BASED VOTING ALGORITHM WITH APPLICATION TO PARSE RERANKING
Authors: Libin Shen and Aravind K. Joshi
Primarily assigned technology terms:
- algorithm
- boosting
- boosting algorithm
- chunking
- classification
- classifier
- classifiers
- computing
- decoding
- dynamic programming
- global optimization
- kernel
- kernels
- lagrange multiplier
- learner
- learning
- learning algorithms
- linear kernel
- local normalization
- local optimization
- loss function
- machine learning
- machine learning algorithms
- mapping function
- markov random fields
- nlp
- normalization
- np chunking
- optimization
- ordinal regression
- parse reranking
- parse scoring
- parser
- parsing
- perceptron
- perceptron algorithm
- pos tagging
- processor
- quadratic programming
- regression
- regression algorithm
- reranking
- risk minimization
- scoring
- scoring function
- search
- searching
- string subsequence kernel
- supertagging
- support vector machines
- svm classifier
- svm-based voting
- tagging
- text chunking
- tree kernel
- viterbi
- viterbi search
- voted perceptron
- voted perceptron algorithm
- voting
- weighting
Other assigned terms:
- approach
- bias
- classification problem
- computational complexity
- context-free grammars
- data set
- derivation
- derivation tree
- elementary tree
- experimental results
- f-score
- feature
- feature set
- feature space
- feature vector
- grammar
- grammars
- history-based model
- hypotheses
- hypothesis
- hypothesis space
- index
- kernel function
- knowledge
- labeling
- learning machine
- lexicalized tree
- linear combination
- log-likelihood
- mapping
- maps
- markov models
- measure
- measures
- method
- n-grams
- nlp applications
- parse
- parse reranking problem
- parse tree
- penn treebank
- penn wsj treebank
- precision
- probabilities
- probability
- relation
- segments
- sentence
- sentences
- sequential model
- sigmoid function
- subtrees
- support vector
- svms
- test phase
- text
- theorem
- theory
- training
- training data
- training dataset
- training phase
- training samples
- tree
- tree adjoining grammar
- treebank
- trees
- wsj treebank
