Posts Tagged ‘Damerau-Levenshtein’

Similarity algorithms

June 28, 2010

I have recently been researching a record linkage techniques, and part of this process I have been reminding myself of certain algorithms, and in other case learning these for the first time. As is my way, I typically try and turn the algorithm into code to allow me to understand and learn it. I have coded up examples of the following algorithms in Java:

Levenshtein
Damerau-Levenshtein
Hamming
Jaro Winkler
N-Gram
Markov Chain

Out of curiosity I decided to publish each one separately so I can see from the stats which is the most popular. In a month or so I will add a histogram of the hits for each one (if there are any!).

I hope this is useful.

Be warned, I offer no warranty or guarantee on this code, any changes / enhancements / corrections / or use of this code should be attributed to itssmee, CHIME, UCL (where I am doing my research) and shared with the community.

Advertisements