Feature-rich sub-lexical language models using a maximum entropy approach for German LVCSR.
M. Ali Basha ShaikAmr El-Desoky MousaRalf SchlüterHermann NeyPublished in: INTERSPEECH (2013)
Keyphrases
- maximum entropy
- language model
- spoken term detection
- language modeling
- context sensitive
- maximum entropy principle
- probabilistic model
- n gram
- document retrieval
- information retrieval
- retrieval model
- speech recognition
- query expansion
- test collection
- conditional random fields
- query terms
- named entity recognition
- language models for information retrieval
- feature vectors
- information retrieval systems
- wordnet
- vector space model
- pseudo relevance feedback
- translation model
- machine learning
- natural language processing
- co occurrence
- higher order