A Comparison of Word-Embeddings in Emotion Detection from Text using BiLSTM, CNN and Self-Attention.
Marco PolignanoPierpaolo BasileMarco de GemmisGiovanni SemeraroPublished in: UMAP (Adjunct Publication) (2019)
Keyphrases
- english words
- sentence level
- string matching
- english text
- text corpus
- text to speech synthesis
- text segments
- text mining
- co occurrence
- keywords
- lexical features
- related words
- information retrieval
- detection algorithm
- printed text
- object detection
- detection method
- text documents
- linguistic information
- text retrieval
- word pairs
- concept space
- training corpus
- sentence similarity
- cellular neural networks
- word sense disambiguation
- printed documents
- convolutional neural network
- visual attention
- natural language text
- text input
- syntactic categories
- word counts
- semantic information
- lexical information
- vector space
- multiword
- complex background