Database Reference
In-Depth Information
The user- and item-factor matrices
These models are often also called latent feature models, as we are trying to discover
some form of hidden features (which are represented by the factor matrices) that account
for the structure of behavior inherent in the user-item rating matrix. While the latent fea-
tures or factors are not directly interpretable, they might, perhaps, represent things such as
the tendency of a user to like movies from a certain director, genre, style, or group of act-
ors, for example.
As we are directly modeling the user-item matrix, the prediction in these models is relat-
ively straightforward: to compute a predicted rating for a user and item, we compute the
vector dot product between the relevant row of the user-factor matrix (that is, the user's
factor vector) and the relevant row of the item-factor matrix (that is, the item's factor vec-
tor).
This is illustrated with the highlighted vectors in the following diagram:
Search WWH ::




Custom Search