EBERT: A lightweight expression-enhanced large-scale pre-trained language model for mathematics education.
Zhiyi DuanHengnian GuYuan KeDongdai ZhouPublished in: Knowl. Based Syst. (2024)
Keyphrases
- lightweight
- language model
- pre trained
- mathematics education
- language modeling
- n gram
- probabilistic model
- speech recognition
- information retrieval
- retrieval model
- query expansion
- mixture model
- training data
- ad hoc information retrieval
- context sensitive
- test collection
- training examples
- smoothing methods
- translation model
- neural network
- control signals
- dirichlet prior
- text classification
- hidden markov models
- pairwise
- data sets