ItemKNN

Reference:

Aiolli,F et al. Efficient top-n recommendation for very large scale binary rated datasets. In Proceedings of the 7th ACM conference on Recommender systems (pp. 273-280). ACM.

class recbole.model.general_recommender.itemknn.ComputeSimilarity(dataMatrix, topk=100, shrink=0, method='item', normalize=True)

Bases: object

compute_similarity(block_size=100)

Compute the similarity for the given dataset

Parameters:

block_size (int) – divide matrix to \(n\_rows \div block\_size\) to calculate cosine_distance if method is ‘user’, otherwise, divide matrix to \(n\_columns \div block\_size\).

Returns:

The similar nodes, if method is ‘user’, the shape is [number of users, neigh_num], else, the shape is [number of items, neigh_num]. scipy.sparse.csr_matrix: sparse matrix W, if method is ‘user’, the shape is [self.n_rows, self.n_rows], else, the shape is [self.n_columns, self.n_columns].

Return type:

list

class recbole.model.general_recommender.itemknn.ItemKNN(config, dataset)

Bases: GeneralRecommender

ItemKNN is a basic model that compute item similarity with the interaction matrix. Adjusting the value of ‘knn_method’ in the config file sets the method to either ItemKNN or UserKNN, respectively.

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(user, item)

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

input_type = 1

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

training: bool

type = 5