Latent Semantic Analysis (LSA)
- is a natural language processing method that uses singular value decomposition to analyze relationships between documents and the terms they contain
- uses a mathematical technique to scan unstructured data to find hidden relationships between terms and concepts
- is a popular dimensionality-reduction technique that reformulates text data in terms of latent features