search

Leavingseason / OpenLearning4DeepRecsys

Some deep learning based recsys for open learning.
398 stars 153 forks source link

ccfnet #13

Closed Edogawa-Konan closed 6 years ago

Edogawa-Konan commented 6 years ago

In your ccfnet 91~92 lines user_embeddings = tf.nn.embedding_lookup_sparse(W_user_attr, user_sp_ids, user_sp_weights, name='user_embeddings', combiner='sum') item_embeddings = tf.nn.embedding_lookup_sparse(W_item_attr, item_sp_ids, item_sp_weights, name='item_embeddings', combiner='sum')` I can't understand the second parameter, user_sp_ids, which is the feature indices(I think). According to my understanding, you use those feature indices to choose rows from W_user_attr. So can you explain why did it ? Thank you!

Leavingseason commented 6 years ago

user_sp_ids are the indices of users' attributes. user_embeddings and item_embeddings are users' and items' latent embeddings learnt from attributes (sparse content features), which is in contrast to CF embedding.

Edogawa-Konan commented 6 years ago

Yeah，user_sp_ids are the indices of user's attributes. The former expression may not be accurate. I fully understand, Sincerely thank you!