Paste two texts to analyse their similarity using TF-IDF vectors, cosine similarity, n-gram overlap, and word-level diff highlighting. See exactly which words and phrases match.
π Text Input
Quick presets:
Document A 0 words
Document B 0 words
or press Ctrl+Enter
π Similarity Score
βsimilarity
β
β
π¬ Visualisations
π Document A
π Document B
word Exact matchword Stem matchword Unique to this docword Stop word
Document A vectorDocument B vectorShared terms (TF-IDF weight)Unique terms
Each axis represents a unique vocabulary term. Documents are represented as vectors where each component is the TF-IDF weight of that term. The angle between vectors corresponds to their cosine similarity. Points near the centre have low TF-IDF weight (common or rare terms).
The cosine similarity measures the angle between the two document vectors. An angle of 0Β° = identical (similarity 1.0). An angle of 90Β° = completely unrelated (similarity 0.0). The length of the vectors does not matter β only the direction.
N-gram size:
Each cell shows whether a given n-gram appears in Document A (left), Document B (right), or both. Red = shared. Blue = A only. Orange = B only.