There's something like automatic completion of Korean in the Naver search box.
If you type "Shoes", I want to expose the recommended search words through automatic completion in the process of 'ㅅ', "Shi", "Shinb", "Shoes".
I'm currently using elasticsearch for my search engine, so it would be better if there was a way to use it in this environment if possible. Is there an elastic search tokenizer that spits out the process of completing Korean typing super-intermediate? Should I make it and use it?
It would be better to have an open source, but if there is no open source, I would like to hear the approach and advice of the implementation.
elasticsearch search
If you use DB instead of elastic search, I think you can handle it as follows.
It is a method of separating Hangul into its own resources, storing it as a separate column, and processing it through LIKE search. For example, in the case of "Shoes", save "ㄴㅂㄹ'ㅏ 을" and separate the letters that entered the search box, and search for "'%" when 'ㅅ' is entered, and search for "'%" when "Si" is entered, the desired result will be obtained.
Please refer to the following for Korean alphabet separation.
© 2024 OneMinuteCode. All rights reserved.