TF-IDF - six documents, preprocessing toggles and full matrix

Data Science & Big Data Analytics • ITec4121 • Wolkite University • Online-First

Six worked documents, a term picker with per-document TF and TF-IDF, and a button that reveals the complete term matrix. Four live preprocessing parameters - lowercasing, length-normalised TF, smoothed IDF and light stemming - plus a top-terms ranking table. The worked case makes it immediate: “analysis” appears in all six documents, so df = 6, IDF = log(6/6) = 0 and its TF-IDF = 0 everywhere, however frequent the term.