Low Frequency Names Exhibit Bias and Overfitting in Contextualizing Language Models.
Robert WolfeAylin CaliskanPublished in: CoRR (2021)
Keyphrases
- language model
- low frequency
- high frequency
- frequency domain
- language modeling
- probabilistic model
- wavelet transform
- speech recognition
- query expansion
- document retrieval
- n gram
- information retrieval
- retrieval model
- discrete wavelet transform
- language modelling
- subband
- test collection
- wavelet coefficients
- statistical language models
- decision trees
- document ranking
- keywords
- low and high frequency
- high resolution
- query terms
- context sensitive
- translation model
- fusion rules
- language models for information retrieval
- query specific
- named entities
- vector space model
- similarity measure
- pseudo relevance feedback
- high frequency components
- multiscale
- okapi bm
- feature extraction
- search engine