ACL RD-TEC 1.0 Summarization of N04-1005
Paper Title:
BALANCING DATA-DRIVEN AND RULE-BASED APPROACHES IN THE CONTEXT OF A MULTIMODAL CONVERSATIONAL SYSTEM
BALANCING DATA-DRIVEN AND RULE-BASED APPROACHES IN THE CONTEXT OF A MULTIMODAL CONVERSATIONAL SYSTEM
Authors: Srinivas Bangalore and Michael Johnston
Primarily assigned technology terms:
- 10-fold crossvalidation
- adaboost
- adaboost classifier
- algorithm
- argument extraction
- automaton
- bootstrapping
- city guide and navigation
- classification
- classification approach
- classifier
- classifiers
- computing
- constraint solver
- conversational speech recognition
- conversational system
- corpus collection
- coupling
- crossvalidation
- data-driven approach
- dialog manager
- encoding
- finite-state machine
- finite-state transducer
- gesture recognition
- graphical interface
- identification
- information gathering
- information seeking
- language model adaptation
- language modeling
- machine translation
- matching
- matching algorithm
- meaning representation
- model adaptation
- model building
- modeling
- multimodal dialog
- multimodal generation
- multimodal integration
- multimodal understanding
- multimodal user interface
- n-gram language modeling
- navigation
- nlp
- parser
- parsing
- pattern matching
- pattern-matching
- phone-based matching
- predicate classification
- rapid prototyping
- recognition
- recognition system
- recognizer
- retrieving
- robust speech recognition
- sampling
- search
- speech recognition
- speech recognition and understanding
- speech recognition system
- speech recognizer
- speech understanding
- speech understanding system
- transducer
- transducers
- user interface
- utterance understanding
- weak classifier
- weighted transducer
- word-based matching
Other assigned terms:
- adjunction
- alphabet
- approach
- asr output
- baseline model
- case
- chunks
- classification error
- classification model
- composition
- concept
- concepts
- conversational speech
- conversational speech corpus
- corpora
- device
- dialog
- dialogue state
- distribution
- domain-specific corpora
- evaluation metrics
- fact
- formalism
- formalisms
- generation
- gesture
- grammar
- grammars
- grammaticality
- hand-crafted grammar
- identification accuracy
- implementation
- interpretation
- interpreter
- language model
- language models
- language variation
- lattice
- lattices
- lexical item
- lexicalized tree-adjoining grammar
- lexicon
- likelihood
- linguistic
- ltag grammar
- meaning
- measure
- method
- multimodal grammar
- multimodal input
- multimodal utterance
- n-best list
- n-gram
- n-gram language model
- names
- out-of-domain corpus
- parse
- posterior
- posterior probability
- predicates
- probabilities
- probability
- procedure
- process
- query
- recognition component
- recognition errors
- representations
- rule sets
- scalability
- sentence
- sentences
- speech corpus
- speech recognition errors
- spoken language
- statistical language model
- supertags
- switchboard corpus
- symbol
- symbols
- syntax
- synthetic speech
- tags
- tagset
- technique
- terminals
- terms
- text
- tokens
- training
- training data
- transcriptions
- tree-adjoining grammar
- trees
- trigram
- trigram language model
- trigram model
- understanding
- understanding component
- usability
- user
- utterance
- verb
- vocabulary
- word
- word corpus
- word sequences
- words
- xtag
