Discussion on Chinese Fuzzy Retrieval in Databases

Now encounter such a scenario, I now have nearly tens of thousands of data in the database.Their format is one Chinese, one English, and one id.
Now start to search Chinese, but his sql is fuzzy search.Should be %x%.How to make the program quickly respond to the retrieved data

Try and discuss:
1. First of all, %x% in mysql; this will directly lead to the failure of the index; everyone should not publish the index; it is impossible to increase the hardware performance; I just want to discuss now,Regarding design, performance, and programming issues
2. Personal thoughts, I have thought about converting Chinese to base64, through Chinese word segmentation, and then matching and searching.
3. There is another way, after checking all the data, put the data in the container, and retrieve the data by writing a small algorithm

Want to ask everyone, if you have experience in this area, you can quickly retrieve data for fuzzy retrieval in Chinese