Need help with DSIN?
Click the “chat” button below for chat support from the developer who created it, or find similar developers for support.

About the developer

312 Stars 107 Forks Apache License 2.0 7 Commits 12 Opened issues


Code for the IJCAI'19 paper "Deep Session Interest Network for Click-Through Rate Prediction"

Services available


Need anything else?

Contributors list

No Data

Deep Session Interest Network for Click-Through Rate Prediction

Experiment code on Advertising Dataset of paper Deep Session Interest Network for Click-Through Rate Prediction(

Yufei Feng , Fuyu Lv, Weichen Shen and Menghan Wang and Fei Sun and Yu Zhu and Keping Yang.

In Proceedings of 28th International Joint Conference on Artificial Intelligence (IJCAI 2019)

Operating environment

please use

pip install -r requirements.txt
to setup the operating environment in

Download dataset and preprocess

Download dataset

  1. Download Dataset Ad Display/Click Data on
  2. Extract the files into the

Data preprocessing

  1. run
    , sample the data by user
  2. run
    , generate historical session sequence for each user

Training and Evaluation

Train DIN model

  1. run
    ,generate input data
  2. run

Train DIEN model

  1. run
    ,generate input data(It may take a long time to sample negative samples.)
  2. run

Train DSIN model

  1. run
    ,generate input data
  2. run
    > The loss of DSIN with
    may be NaN sometimes on Advertising Dataset and it remains a confusing problem since it never occurs in the production environment.We will work on it and also appreciate your help.


This project is licensed under the terms of the Apache-2 license. See LICENSE for additional details.

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.