Combating the Curse of Multilinguality in Cross-Lingual WSD by Aligning Sparse Contextualized Word Representations.
Gábor BerendPublished in: NAACL-HLT (2022)
Keyphrases
- cross lingual
- word sense
- word sense disambiguation
- machine translation
- high dimensional
- translation model
- target word
- parallel corpus
- word segmentation
- language modeling
- language independent
- cross lingual information retrieval
- word alignment
- statistical machine translation
- cross language
- machine translation system
- text classification
- parallel corpora
- n gram
- natural language processing
- bilingual dictionaries
- indian languages
- language model
- target language
- knowledge representation
- document clustering
- probabilistic topic models
- co occurrence
- web pages
- word pairs
- reinforcement learning
- natural language
- part of speech
- probabilistic model
- high dimensional data
- low dimensional
- dimensionality reduction