/SVD-and-Truncated-SVD

Primary LanguageJupyter Notebook

SVD-and-Truncated-SVD

  1. Calculate the tf-idf feature vectors from doc file to obtain the document-to-term matrix
  2. Apply truncated SVD to the document-to-term matrix with k=2 to obtain the representation of reduced dimension for each of the documents
  3. Convert the query 𝒒={attract, TV, program} to its representation of the reduced dimension
  4. Rank the four documents accordingly