Class ShingleBased

java.lang.Object
info.debatty.java.stringsimilarity.ShingleBased
Direct Known Subclasses:
Cosine, Jaccard, QGram, SorensenDice

@Immutable public abstract class ShingleBased extends Object
Abstract class for string similarities that rely on set operations (like cosine similarity or jaccard index). k-shingling is the operation of transforming a string (or text document) into a set of n-grams, which can be used to measure the similarity between two strings or documents. Generally speaking, a k-gram is any sequence of k tokens. We use here the definition from Leskovec, Rajaraman & Ullman (2014), "Mining of Massive Datasets", Cambridge University Press: Multiple subsequent spaces are replaced by a single space, and a k-gram is a sequence of k characters. Default value of k is 3. A good rule of thumb is to imagine that there are only 20 characters and estimate the number of k-shingles as 20^k. For small documents like e-mails, k = 5 is a recommended value. For large documents, such as research articles, k = 9 is considered a safe choice.
  • Field Details

    • DEFAULT_K

      private static final int DEFAULT_K
      See Also:
    • k

      private final int k
    • SPACE_REG

      private static final Pattern SPACE_REG
      Pattern for finding multiple following spaces.
  • Constructor Details

    • ShingleBased

      public ShingleBased(int k)
      Parameters:
      k -
      Throws:
      IllegalArgumentException - if k is <= 0
    • ShingleBased

      ShingleBased()
  • Method Details

    • getK

      public final int getK()
      Return k, the length of k-shingles (aka n-grams).
      Returns:
      The length of k-shingles.
    • getProfile

      public final Map<String,Integer> getProfile(String string)
      Compute and return the profile of s, as defined by Ukkonen "Approximate string-matching with q-grams and maximal matches". https://www.cs.helsinki.fi/u/ukkonen/TCS92.pdf The profile is the number of occurrences of k-shingles, and is used to compute q-gram similarity, Jaccard index, etc. Pay attention: the memory requirement of the profile can be up to k * size of the string
      Parameters:
      string -
      Returns:
      the profile of this string, as an unmodifiable Map