mitao-cat / feature-level_bias

feature_level_bias

This is the implementation of our paper in WWW'24 (short):

Understanding and Counteracting Feature-Level Bias in Click-Through Rate Prediction

By Jinqiu Jin, Sihao Ding, Wenjie Wang, and Fuli Feng

You can also refer to the long paper version by this link.

1. To reproduce results in normal biased tests:

   cd ./code
   python reduction.py --dataset ml-1m --model nfm --alpha 0
   # dataset: {'ml-1m','book','kuairand'}
   # model: {'fm', 'nfm'}
   # alpha: 1->basemodel; 0->reduction

2. To reproduce results in debiased tests:

   cd ./code
   # model: {'fm', 'nfm'}
   # reconstruction: 0->basemodel; 1->reconstruction
   python reconstruction.py --reconstruction 1 --model nfm

Reference: shenweichen/DeepCTR-Torch - GitHub