AI for recommender system

I am going to start AI recommender system(Thai language). I have some question to ask…

  1. Ways to tokenization efficiently. Currently the method I started from create model for tokenize the text, create dict to remove the word that wrong, duplicate word, and custom some word to tokenize. I think this is take so much time I’m not sure is this efficient ways to do so…
  2. I plan to use content based model to recommend. Any recommendation for the model ?

Appreciated for your answer and response in advance

There are a few ready to use tokenizers and you should search the web. Read also this articles maybe you can use the one described here:

1 Like

Great! I will try to implement this with my data. Thank you so much.

1 Like