ACL RD-TEC 1.0 Summarization of H05-1115
Paper Title:
USING RANDOM WALKS FOR QUESTION-FOCUSED SENTENCE RETRIEVAL
USING RANDOM WALKS FOR QUESTION-FOCUSED SENTENCE RETRIEVAL
Authors: Jahna Otterbacher and Gunes Erkan and Dragomir Radev
Primarily assigned technology terms:
- algorithm
- answer extraction
- answer selection
- chronological ordering
- computational linguistics
- document re-ranking
- extraction system
- google search engine
- graph representation
- graph-based centrality
- graph-based method
- human language
- human language technology
- identification
- iterative algorithm
- language processing
- language technology
- multi-document summarization
- natural language processing
- news summarization
- nlp
- normalization
- passage retrieval
- processing
- query-based sentence retrieval
- question answering
- question answering systems
- random walk
- ranking
- re-ranking
- retrieval system
- scoring
- search
- search engine
- sentence retrieval
- sentence selection
- snippet retrieval
- summarization
- summarization system
Other assigned terms:
- annotator
- annotators
- approach
- association for computational linguistics
- baseline model
- bias
- break
- case
- cluster
- clusters
- concept
- content words
- context information
- cosine measure
- cosine similarity
- data set
- data sets
- distribution
- document
- document cluster
- document rank
- document set
- eigenvector centrality
- evaluation metrics
- evaluations
- fact
- formal description
- gold standard
- human judgment
- idf-weighed word
- index
- intelligence
- inter-sentence similarity
- interpretation
- kappa
- linguistics
- markov chain
- measure
- method
- natural language
- pagerank
- paragraphs
- parameter space
- paraphrases
- passage
- probability
- process
- query
- russian
- sentence
- sentences
- similarity scores
- similarity threshold
- stem
- system performance
- technology
- terms
- test data
- test data set
- text
- textual units
- theory
- training
- training data
- training phase
- transition probability
- understanding
- user
- web pages
- word
- words
