ACL RD-TEC 1.0 Summarization of W03-1209
Paper Title:
STATISTICAL QA - CLASSIFIER VS. RE-RANKER: WHAT'S THE DIFFERENCE?
STATISTICAL QA - CLASSIFIER VS. RE-RANKER: WHAT'S THE DIFFERENCE?
Authors: Deepak Ravichandran and Eduard Hovy and Franz Josef Och
Primarily assigned technology terms:
- answer extraction
- approximation
- classification
- classification approach
- classifier
- factoid question answering
- generalized iterative scaling
- identification
- information retrieval
- ir system
- iterative scaling
- likelihood training
- maximum entropy
- maximum entropy model
- maximum entropy qa
- maximum likelihood
- maximum likelihood training
- modeling
- post-processing
- pre-processing
- qa system
- question answering
- re-ranker
- re-ranking
- rule-based system
- search
- search engine
- segmentor
- statistical answer pinpointing
- statistical methods
- statistical qa
- statistical system
- tagging
- validation
- web search
- web search engine
Other assigned terms:
- advanced question answering
- answer type
- approach
- baseline performance
- candidate answer
- case
- chunk
- chunks
- classification problem
- community
- data set
- data sets
- decision rule
- distribution
- entropy
- entropy formulation
- entropy models
- evaluation function
- evaluation metric
- evaluations
- events
- factoid question
- feature
- feature set
- identification module
- intelligence
- likelihood
- maximum entropy models
- measure
- method
- model parameters
- natural language
- nist
- oracle
- parse
- parse tree
- phrase
- precision
- precision measure
- probabilities
- probability
- probability distribution
- process
- query
- question-answer pair
- research and development
- segments
- sentence
- sentences
- statistical framework
- stem
- system architecture
- technique
- training
- training corpus
- training criterion
- training data
- training phase
- tree
- word
- words
