search

guoyang9 / NFM-pyorch

A pytorch implementation for He et al. Neural Factorization Machines for Sparse Predictive Analytics on SIGIR 2017.
38 stars 11 forks source link

Dataset format #1

Closed AmazingDD closed 5 years ago

AmazingDD commented 5 years ago

Can u tell me what is the exact meaning of each column in ml-tag.train.libfm file?

guoyang9 commented 5 years ago

This code is mainly designed for tackling the implicit feedback problem in recommendation. To this end, the rating is normalized into {-1, 1}, where -1 denotes there is no interaction between the user and the item, while 1 denotes opposite.

Therefore, the first value is the ground-truth rating, following by a series of feature id (before semicolon) and feature value (after semicolon).