[ASK] Models with item vector features

[mmosc]

Hi!

First of all thanks for the framework :) I would like to use it to benchmark different models allowing the use item vector features (i.e. FeatureModality).

By following this tutorial I was able to train a TextModality-based model with a FeatureModality, but by looking at the code I have the impression that the feature vectors are not taken "as they are" but further processed as text, i.e. by computing a Bag of Words representation of the features, and using this BoW as input to the model, which is not what I want. Is there a way to use FeatureModality for models that use GraphModality, TextModality, ImageModality, or SentimentModality, leaving the FeatureModality vectors as they are, without further pre-processing happening under the hood?

Alternatively: Is there a list of all the models implemented in cornac that leverage item vector features as FeatureModality (in the same way as BiVAECF, for instance)?

Cheers! Marta