TF IDF Weight Calculator

Work out tf idf weight instantly with clear inputs, formula shown and shareable results.

About TF IDF Weight Calculator

TF-IDF multiplies how often a term appears in a document by how rare it is across the corpus. The rarity factor ln(N/df) is what suppresses stopwords: a word in every document has an IDF of zero and therefore no weight, while a term in one percent of documents gets a large multiplier and dominates the vector.

Formula

TF-IDF

tf = count / document length; idf = ln(N / df); weight = tf x idf

N: Documents in the corpus
df: Documents containing the term

Worked example

With the default values below, TF IDF Weight Calculator returns TF-IDF weight of 0.08846. Change any input above and every figure here is recalculated in your browser.

Inputs used

Occurrences of the term in the documentminimum 0: 8
Words in the documentminimum 1: 400
Documents in the corpusminimum 1: 10,000
Documents containing the termminimum 1: 120

Results

TF-IDF weight: 0.08846
Normalised term frequency: 0.02000
Inverse document frequency: 4.42285
Share of documents containing the term: 1.200%

Step by step

1Values used
2TF-IDF
3TF-IDF weight= 0.08846
4Normalised term frequency= 0.02000
5Inverse document frequency= 4.42285
6Share of documents containing the term= 1.200

Assumptions and limits

The result is derived only from the 4 inputs listed above. Anything not among them is not part of the calculation.
In particular, this calculator does not take inflation, tax or fees or charges into account.
Figures are rounded for display to 5 decimal places, so totals built from several rounded parts can differ slightly from the unrounded arithmetic.

How these calculators are built and tested is described on the Methodology page. If a figure looks wrong, please report it.

Frequently Asked Questions

Why is my TF-IDF zero?

Because the term appears in every document, making ln(N/df) = ln(1) = 0. Smoothed variants add 1 to the denominator and the result to avoid this.

How does BM25 improve on TF-IDF?

BM25 saturates term frequency so the twentieth occurrence adds little, and normalises by document length relative to the corpus average, which makes it far more robust on mixed-length documents.