Concatenated p-mean Word Embeddings as Universal Cross-Lingual Sentence Representations.
Andreas RückléSteffen EgerMaxime PeyrardIryna GurevychPublished in: CoRR (2018)
Keyphrases
- cross lingual
- parallel corpus
- sentence level
- translation model
- sentiment classification
- source language
- machine translation
- language modeling
- word alignment
- language independent
- cross lingual information retrieval
- cross language
- word sense
- text classification
- word segmentation
- cross language information retrieval
- statistical machine translation
- target language
- machine translation system
- query translation
- parallel corpora
- language model
- natural language
- n gram
- vector space
- bilingual dictionaries
- feature vectors
- noun phrases
- news articles
- sentiment analysis
- syntactic information
- indian languages
- question answering
- learning algorithm
- co occurrence