Exploiting Twitter as Source of Large Corpora of Weakly Similar Pairs for Semantic Sentence Embeddings.
Marco Di GiovanniMarco BrambillaPublished in: EMNLP (1) (2021)
Keyphrases
- semantically similar
- natural language
- dependency relations
- social media
- social networks
- word frequency
- semantic content
- online social networks
- recognizing textual entailment
- dimensionality reduction
- semantic roles
- natural language processing
- syntactic structures
- semantic web
- semantic information
- syntactic dependencies
- semantic similarity
- text representation
- helmholtz principle
- twitter users
- user generated content
- social networking
- low dimensional
- domain specific