How can I optimize my cosine similairty function that compares row values between two columns?

Question

I am attempting to calculate the cosine similarity between a title and search query that are stored within a pandas dataframe, but am struggling to find the optimal method.