Similitude Attentive Relation Network for Click-Through Rate Prediction

This is the code for Similitude Attentive Relation Network forClick-Through Rate Prediction. It is an example for Amazon Electronics dataset and you can try other datasets easily.

prepare data

split_by_user

sh prepare_data_split_by_user_electronics.sh

train model

python script/train.py data/Electronics_split_by_user/

Requirements:

Python 2.7

Tensorflow 1.4

Note: the code is heavily dependent on DIEN code in the year of 2018.