Since Moniker made the dataset open on GitHub you're free to try your luck! Above link is a pretty good implementation.