How to calculate the maximum cosine similarity between a user's reviews in spark? (python)

Question

I have a spark data frame with one column is user_id, another column is review_text.

I want to return the third column "maximum_cosine_similarity", which is