This repository is originated from luongvuongnquyen/DaGzang.

A Synthetic Data Generator for Cross-domain Recommendation Services

This study aims to propose a platform, called DaGzang, to generate synthetic datasets provide for cross-domain recommendation systems. DaGzang platform working according to the complete loop and it has three major roles as follows (i) detect the overlap association (data distribution pattern) between the real-world datasets, (ii) generate synthetic datasets based on these real-world datasets, and (iii) evaluate the quality of the generated synthetic datasets. Additionally, DaGzang is deployed into DakGalBi, our cross-domain recommender system, in order to validate the usefulness of the synthetic datasets generated.

The codebase is implemented in Java (develop kit 8.0) | Apache Tomcat (64-bit). Package versions used for development are just below.

IDE Netbean 12
MySQL 8.0 
Java 8.0
Apache Tomcat 8
Spring Framework
Maven

1. Start web sever

$mvnw tomcat8:run

2. Create WAR file

$mvnw clean package

3. Start MySQL

docker container run -d -p 3306:3306  \
-e MYSQL_ROOT_PASSWORD=password \
-e MYSQL_DATABASE=dagzang \
-e MYSQL_USER=root \
-e MYSQL_PASSWORD=ke@cau \
--name dagzang mysql:8.0.21

4. Run in browser

• http://localhost:8084/dakgalbi
• Try online version: http://recsys.cau.ac.kr:8084/dakgalbi

In case of queries, please email: nguyenluongvuong@gmail.com