based it off this 2017 dataset from UCSD (https://cseweb.ucsd.edu/~jmcauley/datasets/goodreads.html)

some fun writing CUDA accelerated code to cluster the data and then find similarity scores