• Chromium
• Mozilla Core
• Mozilla Firefox
• NetBeans
• eclipse

Note: The code works with python 3.6, tensorflow 1.14 and keras 2.3.1.

1. You need download data sets mentioned above first.
2. For Chromium, Mozilla Core and Mozilla Firefox, you can directly use after upziping under data folder.
3. For NetBeans and eclipse, you need run the SQL file in MySQL. Then big_id, short_desc, assigned_to, bug_status and resolution field in table bugs and bug_id, bug_when and thetext field in table longdescs are extracted in a json format. Note that the format of bug_when field is set to a sortable style like yymmdd 00-00-00. After that you can put the file under data folder.
4. Run HAN-XXX.ipynb file in Jupyter Notebook.
5. GloVe word vectors need to be trained separately, cleck here for concrete manner.