Independent Components of Word Embeddings Represent Semantic Features.
Tomás MusilDavid MarecekPublished in: CoRR (2022)
Keyphrases
- semantic features
- independent components
- independent component analysis
- visual features
- semantic information
- wordnet
- text classification
- document clustering
- factor analysis
- semantic similarity
- co occurrence
- n gram
- low level features
- blind source separation
- keywords
- dimensionality reduction
- word sense disambiguation
- feature vectors
- syntactic features
- data mining
- principal components
- text categorization
- principal component analysis
- knn
- xml documents
- feature space
- object recognition
- computer vision
- machine learning