Improving Relevance Feedback in Language Modeling Approach: Maximum a Posteriori Probability Criterion and Three-Component Mixture Model.
Seung-Hoon NaIn-Su KangJong-Hyeok LeePublished in: IJCNLP (2004)
Keyphrases
- mixture model
- language modeling
- language model
- maximum a posteriori probability
- retrieval model
- relevance feedback
- query expansion
- document retrieval
- probabilistic model
- pseudo relevance feedback
- markov random field
- information retrieval
- gaussian mixture model
- test collection
- n gram
- dirichlet prior
- retrieval effectiveness
- map estimation
- em algorithm
- generative model
- probability density function
- vector space model
- image retrieval
- smoothing methods
- retrieved documents
- document length
- active learning
- expectation maximization
- feature selection
- information retrieval systems
- text classification
- maximum likelihood
- bayesian networks
- relevant documents
- natural language processing
- graphical models
- search engine
- unsupervised learning
- model selection
- learning algorithm
- machine learning