Now we can make the best recommendation for our user, based on their user feature vector and the features of the unrated unseen movies in our database.

We’ll compute the dot product to measure the similarity between our user and all the remaining unranked movies in our database.

The movie with the greatest similarity measure is our top recommendation for our user.

The dot product is found by taking the component wise product across each dimension and adding the results.

That is, we multiply the user feature vector component wise with the movie feature vector for each movie.

and then sum row wise to compute the dot product.

This gives us the dot product similarity between our user and each of the four movies.

We’ll use these values to make our recommendations.

because star wars has the greatest similarity measure, that will be our top recommendation, followed by the incredibles and then the dark knight and lastly memento.