This is replementation of DMR using DL framework PaddlePaddle.
This repo is modified based on original tensorflow version.
Deep Match to Rank Model for Personalized Click-Through Rate Prediction
This code for DMR is implemented based on the Alimama Dataset (https://tianchi.aliyun.com/dataset/dataDetail?dataId=56), references the codes of DIEN (https://github.com/mouna99/dien).
"alimama_sampled.txt" is a sampled dataset for testing the code.
Notice that, the user behavior data in this dataset contains only category, brand and behavior type, but no item.
run "export LD_LIBRARY_PATH='/usr/local/cuda-10.0/lib64/'" before import paddle