ACL RD-TEC 1.0 Summarization of W04-2905
Paper Title:
USING SOUNDEX CODES FOR INDEXING NAMES IN ASR DOCUMENTS
USING SOUNDEX CODES FOR INDEXING NAMES IN ASR DOCUMENTS
Authors: Hema Raghavan and James Allan
Primarily assigned technology terms:
- approximate string matching
- asr system
- capitalization
- computing
- coreference resolution
- database
- detection and tracking
- document coreference resolution
- document retrieval
- entity normalization
- entity recognition
- entity recognizer
- entity tagging
- grouping
- hidden markov
- hidden markov model
- indexing
- information retrieval
- intelligent information retrieval
- levenshtein
- link detection
- machine translation
- machine translation system
- machine translation systems
- markov model
- matching
- matching technique
- named entity recognition
- named entity recognizer
- named entity tagging
- normalization
- recognition
- recognizer
- retrieving
- sampling
- speech recognizer
- spelling
- spoken document retrieval
- story link detection
- string matching
- tagging
- topic detection
- topic detection and tracking
- translation system
- translation systems
- transliteration
Other assigned terms:
- alphabet
- approach
- asr output
- broadcast news
- canonical form
- case
- confidence score
- confidence scores
- cosine similarity
- cross document coreference
- database record
- detection task
- document
- document coreference
- edit distance
- english vocabulary
- error rate
- fact
- french
- grammatical structure
- heuristics
- index
- levenshtein distance
- lexicon
- measure
- mechanisms
- method
- named entities
- named entity
- names
- opinions
- person names
- phonemes
- precision
- proper names
- punctuation
- queries
- query
- racing
- sentence
- sentence boundaries
- sentence structure
- similarity metric
- similarity metrics
- similarity scores
- stem
- tdt corpus
- technique
- term
- terms
- test set
- text
- topics
- transcripts
- vocabulary
- word
- word error rate
- words
