Clustering the dataset about dkgr HOT 2 CLOSED

Hey, thanks for your interest in our work.

The detailed steps for data pre-processing are:

1. Use TransE (or other pretrained methods) to generate KG embeddings.
2. Use the "pretraining.py" to obtain the mapping from entity node to cluster node.
3. Move the "entity2clusterid.txt" file into the target dataset folder "dataset/{dataset_name}/", e.g., "dataset/nell/".
4. Run the "code/data/preprocessing_scripts/create_vocab.py"

Hope the above information is helpful to you.

from dkgr.

Thank you so much, I think my problem was with the embeddings as I was getting the same results 😂

from dkgr.