Exploration of the Impact of Maximum Entropy in Recurrent Neural Network Language Models for Code-Switching Speech.
Ngoc Thang VuTanja SchultzPublished in: CodeSwitch@EMNLP (2014)
Keyphrases
- maximum entropy
- language model
- recurrent neural networks
- speech recognition
- word error rate
- language modeling
- probabilistic model
- n gram
- feed forward
- neural network
- spoken term detection
- recurrent networks
- document retrieval
- maximum entropy principle
- conditional random fields
- retrieval model
- speech signal
- statistical language models
- test collection
- information retrieval
- smoothing methods
- mixture model
- query expansion
- language modelling
- automatic speech recognition
- query terms
- named entity recognition
- relevance model
- pseudo relevance feedback
- language models for information retrieval
- artificial neural networks
- translation model
- vector space model
- cross lingual
- information retrieval systems
- data clustering
- machine learning