S3Rec

Reference:

Kun Zhou and Hui Wang et al. “S^3-Rec: Self-Supervised Learning for Sequential Recommendation with Mutual Information Maximization” In CIKM 2020.

Reference code:

https://github.com/RUCAIBox/CIKM2020-S3Rec

class recbole.model.sequential_recommender.s3rec.S3Rec(config, dataset)

Bases: recbole.model.abstract_recommender.SequentialRecommender

S3Rec is the first work to incorporate self-supervised learning in sequential recommendation.

Note

Under this framework, we need reconstruct the pretraining data, which would affect the pre-training speed.

calculate_loss(interaction)

Calculate the training loss for a batch data.

Parameters

interaction (Interaction) – Interaction class of the batch.

Returns

Training loss, shape: []

Return type

torch.Tensor

forward(item_seq, bidirectional=True)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

full_sort_predict(interaction)

full sort prediction function. Given users, calculate the scores between users and all candidate items.

Parameters

interaction (Interaction) – Interaction class of the batch.

Returns

Predicted scores for given users and all candidate items, shape: [n_batch_users * n_candidate_items]

Return type

torch.Tensor

get_attention_mask(sequence, bidirectional=True)

In the pre-training stage, we generate bidirectional attention mask for multi-head attention.

In the fine-tuning stage, we generate left-to-right uni-directional attention mask for multi-head attention.

predict(interaction)

Predict the scores between users and items.

Parameters

interaction (Interaction) – Interaction class of the batch.

Returns

Predicted scores for given users and items, shape: [batch_size]

Return type

torch.Tensor

pretrain(features, masked_item_sequence, pos_items, neg_items, masked_segment_sequence, pos_segment, neg_segment)

Pretrain out model using four pre-training tasks:

1. Associated Attribute Prediction

2. Masked Item Prediction

3. Masked Attribute Prediction

4. Segment Prediction

reconstruct_pretrain_data(item_seq, item_seq_len)

Generate pre-training data for the pre-training stage.

training: bool