ACL RD-TEC 1.0 Summarization of P03-1030
Paper Title:
OPTIMIZING STORY LINK DETECTION IS NOT EQUIVALENT TO OPTIMIZING NEW EVENT DETECTION
OPTIMIZING STORY LINK DETECTION IS NOT EQUIVALENT TO OPTIMIZING NEW EVENT DETECTION
Authors: Ayman Farahat and Francine Chen and Thorsten Brants
Primarily assigned technology terms:
- anaphora resolution
- asr system
- automatic speech recognition
- detection and tracking
- event detection
- first story detection
- information detection
- information retrieval
- information retrieval task
- java
- link detection
- link detection system
- matching
- ned detection
- normalization
- part of speech tagging
- part-of-speech tagging
- performance enhancing
- pos tagging
- pre-processing
- preprocessing
- processing
- recall enhancing
- recognition
- segmentation
- similarity calculation
- speech recognition
- speech tagging
- story detection
- story link detection
- story segmentation
- summarization
- tagging
- term weighting
- text segmentation
- topic detection
- topic detection and tracking
- topic tracking
- weighting
Other assigned terms:
- abbreviations
- adjective
- anaphora
- anchors
- approach
- asr stop-list
- broadcast news
- case
- cluster
- clusters
- conditional probabilities
- confidence scores
- corpora
- correlation
- cosine distance
- cosine similarity
- data set
- data sets
- density function
- device
- distribution
- document
- document frequency
- events
- experimental results
- hypothesis
- implementation
- inverse document frequency
- kullback-leibler divergence
- labeled training data
- labeling
- measure
- measures
- method
- names
- noise
- nouns
- null hypothesis
- parallel corpus
- parameter settings
- part of speech
- part-of-speech
- part-of-speech information
- part-of-speech tags
- precision
- priori
- probabilities
- probability
- probability distribution
- query
- recognition errors
- retrieval task
- senses of a word
- server
- similarity measure
- similarity measures
- similarity metrics
- similarity scores
- speech recognition errors
- statistical approach
- statistics
- stem
- stems
- system performance
- tags
- tdt corpus
- technologies
- technology
- term
- terms
- text
- tokens
- topics
- training
- training corpus
- training data
- training set
- verb
- vocabulary
- word
- words
