Near-Duplicate-Detection

• SimHash

  Fingerprinting algorithm preserving cosine similarity. Maps similar documents to nearby hashes; enables efficient near-duplicate detection.

• Shingling

  Shingling represents a document as its set of overlapping n-grams (shingles), enabling near-duplicate detection via Jaccard similarity or MinHash approximations.

• Shingle

  A shingle is an n-gram treated as a set element for document comparison. The term signals a shift from positional sequence analysis to set-based similarity measurement.