There is a base in which letters of other people are recorded. When a new letter appears, you need to find out whether it is similar to the others present in the database and how many percent match. Letters are in the field type TEXT. As I understand it, you can somehow through MATCH, but all the examples are only for finding one phrase in the text, but I need to compare the whole text with other texts. Please share your ideas.

  • Maybe SOUNDEX ()? - cyadvert
  • @cyadvert Certainly not a soundex, it gives out some level of consonance. But this level is rigidly tied to the first letter of the word and in general it only makes sense when working with one word - Mike
  • @ Alexey A match is designed to search for a single phrase, which is why it is a full-text search . I think in mysql there are no ready-made tools for solving your problem. It is generally non-trivial. - Mike
  • not necessarily using mysql tools — php can be, if only you know how - Alexey
  • Well, yes, with SOUNDEX (), I didn’t think it blurted out ... If you can use PHP tools, then you can create an array of source text words (words are array keys; how many times you meet values) and compare it with the same array of other texts .. array_key_intersect can help a lot ... Like an idea, people? - cyadvert

0