/class-similarity

The code to calculate similarity of different dataset

Primary LanguagePython

Class-similarity

This repo offers the code to calculate similarity of different dataset based on dataset labels using cosine similarity and GloVe embedding.

cosine_similarity.py provides the function to calculate the cosine similarity between two GloVe vector.

embedding.py converts labels of dataset to GloVe vectors.

get_avelabel.py gets labels from label json files.

label_similarity.py is the main function to calculate the similarity given the label files.