Dimensionality Reduction

• text data is often high-dimensional, sparse and complex
• computational requirements increase exponentially with the number of dimensions
• it is hence necessary to condense this data into a more usable form
• vector dimensionality reduction in Natural Language Processing (NLP) is the process of reducing the number of features (dimensions) in a vector representation of text data

Dimensionality Reduction techniques include:

• Principal Component Analysis (PCA)
• Singular Value Decomposition (SVD)

Shorter vectors are easier to learn from, and the models that take them are listed in Text Classification.