Maximizing diversity over clustered data

This repo hosts the code for the paper "G. Zhang AND A. Gionis, Maximizing diversity over clustered data, 2020".

The main algorithms are in intra.py, and metrics are run by Jupyter notebooks that start with metric-. The repo adopts sacred and incense to manage experiment results. Therefore you will need a Mongo DB server to run experiments.

External realistic data can be downloaded from links below:

• https://grouplens.org/datasets/movielens/
• https://u.cs.biu.ac.il/~koppel/BlogCorpus.htm
• https://www.aminer.cn/data#Academic-Social-Network