are you aware worthwhile calculations that match two strings after which return a portion in the number of percent individuals two strings match?
And therefore are there some, that actually work with databases too?
The Levenstein distance is really a stride. It essentially informs you the number of figures have to be edited, erased or added, to achieve with a home first towards the second string. I am unsure whether some database systems support that.
However I know without a doubt that an infinitely more simplified formula named Soundex is supported in certain database systems.
It is dependent upon your criteria for similarity. Others have previously known you to definitely Levenstein distance (edit distance is identical factor). That's usually very good, and certainly more language-independent than something similar to soundex. However, remember that Levenstein difference doesn't handle transposition perfectly. Thus:
Levenstein("copy", "cpoy") == 2
If you are trying to cope with human input, transpositions are not unheard of. Whether this is a problem or otherwise is dependent in your metrics for similarity.
It has been some time, however i believe Postgresql has
levenstein() either built-in or available like a
contrib C module.
I believe the issue you are searching for is known as Edit Distance. It's costly to compute generally, but when you're searching for strings within small edit distance of other strings, it's not so bad. There's more details within the Wikipedia article.
Maybe that one is going to be helpful for ya:
Would this help you? I simply went in it. http://stackoverflow.com/questions/188476/comparing-two-strings-producing-a-numeric-delta
The proper way to match two strings? Ask them to venture out for coffee, and when they hit them back, dinner along with a movie. Or possibly they might perform some peer programming? It is dependent around the strings, really. Even coffee can frequently be tricky.