In PROC TEXTMINE, which step in the text mining pipeline transforms raw text documents into a term-by-document matrix?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — you need text parsing using the PARSE statement because it tokenizes and stems raw text, removes stop words, and builds the term-by-document matrix. The other steps work after that matrix exists.
Full Explanation
Text parsing using the PARSE statement is correct because in PROC TEXTMINE, the PARSE statement is responsible for parsing raw text: it tokenizes documents into terms, performs stemming, removes stop words, and builds the term-by-document frequency matrix that serves as input for subsequent analyses. Singular Value Decomposition (SVD) is a dimensionality reduction step applied after the term matrix is built. Sentiment scoring using the SENTIMENT statement scores documents after parsing. Topic modeling using the LDA algorithm discovers topics from the term matrix after parsing.