NGCF

class dhg.models.NGCF(*args, **kwargs)[source]

Bases: torch.nn.Module

The NGCF model proposed in Neural Graph Collaborative Filtering paper (SIGIR 2019).

Note

The user and item embeddings and trainable parameters are initialized with xavier_uniform distribution.

Parameters

num_users (int) – The Number of users.
num_items (int) – The Number of items.
emb_dim (int) – Embedding dimension.
num_layers (int) – The Number of layers. Defaults to 3.
drop_rate (float) – The dropout probability. Defaults to 0.5.

forward(ui_bigraph)[source]

The forward function.

Parameters: ui_bigraph (dhg.BiGraph) – The user-item bipartite graph.