anuradha1992 / heal

Code and the dataset for HEAL: A Knowledge Graph for Distress Management Conversations

To run the python scripts from the beginning, the Reddit dialogues dataset curated from a caefully chosen 8 psychological distress related subreddits, should be requested from the authors and be included inside the folder ./original. This dataset will not be publicly available due to ethical reasons.

The data files "dump", "indices_dump", and "cluster_dump" required to run the scripts and that are stored in pickle format can be downloaded from: https://drive.google.com/drive/folders/1h_e1Qv08yZNYc04S-OXWAr1unn1MEVpI?usp=sharing

heal's Issues

Where can I find the appendices mentioned in the paper?

Dear authors,

Where can I find the appendices mentioned in the paper? There is no appendix at the end of the paper.

Looking forward to your reply!
Thanks a lot!

What script are you referring to in the README?

I'm trying to reproduce your code, What script are you referring to in the README?

Looking forward to your reply. Thanks a lot!

How to get dump.pkl is not covered in the code.

paraphrase-MiniLM-L6-v2 can only maps sentences & paragraphs to a 384 dimensional dense vector space , Why does the embedding in dump.pkl have 768 dimensions.