Imputing Out-of-Vocabulary Embeddings with LOVE Makes Language Models Robust with Little Cost.
Lihu ChenGaël VaroquauxFabian M. SuchanekPublished in: CoRR (2022)
Keyphrases
- language model
- out of vocabulary
- n gram
- language modeling
- probabilistic model
- spoken term detection
- document retrieval
- speech recognition
- query terms
- information retrieval
- query expansion
- test collection
- context sensitive
- retrieval model
- vector space model
- cross language information retrieval
- word segmentation
- named entity recognition
- document representation
- pseudo relevance feedback
- co occurrence
- relevance model
- broadcast news
- knn
- feature selection