From the course: Building Recommender Systems with Machine Learning and AI

Unlock this course with a free trial

Join today to access over 22,400 courses taught by industry experts.

More emerging tech to watch

More emerging tech to watch

- [Instructor] While we are still operating under our bleeding edge alert, let's briefly cover a few more recent research topics that are worth keeping an eye on. Some people have applied a neural network model called word2vec, to recommender problems. If you have text data, to work with, it can be an interesting tool. The idea is you start with some words, that represent the context, or user history you want to use to inform your recommendations. But to make it easier to understand let's just start with the problem of completing someones sentence. We can encode all of the words in that sentence. Feed it through an embedding layer and a hidden layer where the learning takes place and run a softmax classifier on the output of the hidden layer to find the probabilities of each word that might be associated with the ones you fed in this input. In this example, we feed in signals associated with the words to boldly go where no one has. And a properly trained word2vec network might…

Contents