Internet Chinese information retrieval using unconstrained Mandarin speech queries based on a client-server architecture and a PAT-tree-based language model.
Lee-Feng ChienSung-Chien LinJenn-Chau HongMing-Chiuan ChenHsin-Min WangJia-Lin ShenKeh-Jiann ChenLin-Shan LeePublished in: ICASSP (1997)
Keyphrases
- language model
- client server architecture
- speech recognition
- information retrieval
- query terms
- language modeling
- client server
- trec test collections
- cross language retrieval
- test collection
- word error rate
- retrieval model
- document retrieval
- query specific
- n gram
- relevance model
- speech signal
- query expansion
- retrieval systems
- web browser
- automatic speech recognition
- spoken document retrieval
- retrieval effectiveness
- vector space model
- probabilistic retrieval models
- probabilistic model
- trec collections
- database server
- ad hoc retrieval
- query processing
- term dependencies
- word segmentation
- query language
- document length
- pseudo relevance feedback
- document collections
- information retrieval systems
- peer to peer
- relevant documents
- file sharing
- user queries
- out of vocabulary
- database
- retrieval strategies
- web search engines
- information extraction
- text mining
- cross language information retrieval
- chinese english
- cross lingual
- bayesian networks
- term weighting
- xml retrieval
- natural language processing
- tf idf
- question answering
- search engine
- databases