/FastEALS

Experiments codes for SIGIR'16 paper "Fast Matrix Factorization for Online Recommendation with Implicit Feedback "

Primary LanguageJava

sigir16-eals

Experiments codes for SIGIR'16 paper "Fast Matrix Factorization for Online Recommendation with Implicit Feedback "

项目说明

该项目是基于何向南的fastEALS算法的应用和修改。

1、algorithms/MF_fastALS_WRMF是对MF_fastALS的扩充,添加了参数设置选择;

2、main/main_XiamiOnline是针对虾米数据的实验环境:

(1)train文件:单个文件,和librec的训练文件格式一致,userId<space>itemId<space>freq
(2)test文件夹:对每个用户的播放记录文件,文件名为<userId>.txt,每行内容为timestamp<\t>itemId