Factorization Item-to-Item matrix

[itsoum]

Hello Ben,

I have an item-to-item matrix, where items are services and the values of matrix are a mean of graph distances for each pair of services. Graph distances are exported from my dataset where I have application graphs which are consisted by above-mentioned services. This distance I will use as implicity rating for the strongness of each relation between of services. Is there some method with implicity library to factorize this item-to-item matrix in order to estimate the missing values in a matrix' = item-to-item?

I ask because in your basic example you use a user-item matrix.

# train the model on a sparse matrix of item/user/confidence weights
model.fit(item_user_data)
 

Thank you, Ilias