Visualization of topic transitions in SNSs through document embedding and dimensionality reduction.
Tiandong XiaoNaoya OdaYosuke OnouePublished in: J. Vis. (2023)
Keyphrases
- dimensionality reduction
- multidimensional scaling
- nonlinear dimensionality reduction
- structure preserving
- document content
- topic discovery
- document set
- graph embedding
- low dimensional
- latent dirichlet allocation
- locality preserving projections
- high dimensional
- principal component analysis
- embedding space
- social media
- neighborhood preserving
- high dimensional data
- manifold learning
- information retrieval systems
- latent topics
- textual content
- topic hierarchy
- document images
- document level
- data representation
- low dimensional spaces
- document collections
- scientific papers
- multi dimensional scaling
- focused crawler
- topic models
- document clustering
- high dimensionality
- related documents
- information retrieval
- principal components
- web documents
- vector space
- feature extraction
- data analysis
- keywords
- news articles
- linear dimensionality reduction
- document corpus
- automatic summarization
- test collection
- feature selection
- pattern recognition
- relevant documents
- document representation
- social networking sites
- document retrieval
- text documents
- retrieval systems
- user queries
- dimensionality reduction methods
- co occurrence
- data points
- feature space
- image processing
- web pages
- single document summarization
- kernel pca