ACL RD-TEC 1.0 Summarization of P06-1026
Paper Title:
LEARNING THE STRUCTURE OF TASK-DRIVEN HUMAN-HUMAN DIALOGS
LEARNING THE STRUCTURE OF TASK-DRIVEN HUMAN-HUMAN DIALOGS
Authors: Srinivas Bangalore and Giuseppe Di Fabbrizio and Amanda Stent
Primarily assigned technology terms:
- algorithm
- automatic dialog modeling
- call routing
- classi cation
- classi er
- clustering
- computational linguistics
- cross validation
- cross-validation
- data collection
- data-driven approach
- decoder
- decoding
- dependency analysis
- dialog management
- dialog manager
- dialog modeling
- dialog system
- dialog systems
- disambiguation
- discriminative learning
- discriminative modeling
- dynamic programming
- dynamic programming search
- greedy decoding
- hidden markov
- hidden markov models
- human-machine dialog
- incremental parser
- incremental processing
- language generation
- language parsing
- language processing
- language processing systems
- language understanding
- learner
- learning
- machine learning
- maxent
- maximum entropy
- modeling
- natural language generation
- natural language processing
- natural language processing systems
- parser
- parsing
- plan recognition
- processing
- recognition
- regular expression
- reinforcement learning
- scoring
- search
- segmentation
- speech recognition
- spoken dialog
- spoken dialog systems
- structural analysis
- structure prediction
- structuring
- subtask labeling
- syntactic annotation
- syntactic parser
- synthesis
- system deployment
- tagger
- taggers
- tagging
- ten-fold cross-validation
- text-to-speech
- text-to-speech synthesis
- top-down parser
- utterance segmentation
- validation
- viterbi
- viterbi search
Other assigned terms:
- adjunction
- annotated corpus
- annotation
- approach
- asr output
- association for computational linguistics
- automated contact
- background noise
- call center
- case
- chunk
- chunks
- classi cation task
- community
- conditional distribution
- context information
- context-free grammars
- contextual features
- contextual information
- corpora
- cue phrases
- customer care
- damsl
- data set
- data sets
- dialog
- dialog act
- dialog structure
- discourse
- discourse markers
- distribution
- entropy
- entropy models
- error rate
- feature
- feature set
- feature sets
- generation
- generative model
- grammar
- grammars
- hypotheses
- information states
- interpretation
- labeling
- language model
- large corpora
- large corpus
- lattice
- linguistic
- linguistic structure
- linguistics
- local context
- maptask corpus
- markov models
- maximum entropy models
- method
- methodology
- n-gram
- n-gram models
- n-grams
- natural language
- noise
- opinion
- parse
- parse tree
- parsing research
- part-of-speech
- partial parse
- partial parse tree
- pauses
- plan tree
- predicate-argument
- predicate-argument structure
- prediction accuracy
- probabilistic context-free grammars
- probabilistic models
- process
- prosodic features
- recognition error rate
- scalability
- segments
- semantic
- sentence
- sentence boundaries
- speech recognition error
- speech repairs
- spoken language
- supertags
- switchboard corpus
- switchboard-damsl
- syntactic features
- tag set
- tagging scheme
- tags
- task-oriented dialog
- technique
- test set
- text
- toolkit
- training
- training data
- training set
- transcribed speech
- tree
- trees
- trigram
- understanding
- user
- utterance
- vocabulary
- well-formedness constraint
- word
- word trigram
- words
