Notice there are no torch-related code about item2vec, I just want to provide a readable item2vec implementation for researchers
The format of input data only need a pd.DataFrame with 3 columns(user, item, rating)
Since the paper claims that Item2Vec discards time and spatial information of interactions, we can simply fit this model like word2vec with some additional preprocessing sampling methods.
To run the code, just enter this code in your terminal
python main.py
This repository does not provide any predict interface because this model is only to get the Item Embedding information after training, then users could do whatever they want flexibly(get rank lists, find KNN items, etc.)
The embedding weight is stored in model.shared_embedding with dimension (item_num * embedding_dim)
you can customize the hyperparameter settings by changing the dictionary variable args in main.py
args = {
'context_window': 2,
'rho': 1e-5,
'batch_size': 256,
'embedding_dim': 100,
'epochs': 20,
'learning_rate': 0.001,
}
Barkan, Oren, and Noam Koenigstein. "Item2vec: neural item embedding for collaborative filtering." 2016 IEEE 26th International Workshop on Machine Learning for Signal Processing (MLSP). IEEE, 2016.