/The-FacT--Sigir2019

The FacT: Taming Latent Factor Models for Explainability with Factorization Tree

Primary LanguagePython

The FacT

This is the implementation for the paper titled "The FacT: Taming Latent Factor Models for Explainability with Factorization Trees". We provide all the source codes for the algorithm.

Reference

Yiyi Tao, Yiling Jia, Nan Wang and Hongning Wang. The FacT: Taming Latent Factor Models for Explainability with Factorization Trees, SIGIR 2019.

Quick Start (For Linux and Mac)

Data format

In ./data/yelp_train.txt

user_id, item_id, rating, [list of feature opinions]

Example:
1, 0, 4, 1 1 2 1
user_id = 1, item_id = 0, rating = 4, rating for feature 1 = 1, rating for feature 2 = 1.

How to run the algorithm

$ cd code
$ python main.py --train_file ../data/yelp_train.txt --test_file ../data/yelp_test.txt --num_dim 20 --max_depth 6

The results will be stored in ./results/