# Copyright 2018 The Cornac Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ============================================================================
import numpy as np
from tqdm.auto import trange
from ..recommender import Recommender
from ..recommender import ANNMixin, MEASURE_DOT
from ...exception import ScoreException
from ...utils import get_rng
from ...utils.init_utils import xavier_uniform
[docs]
class CDL(Recommender, ANNMixin):
"""Collaborative Deep Learning.
Parameters
----------
name: string, default: 'CDL'
The name of the recommender model.
k: int, optional, default: 50
The dimension of the latent factors.
max_iter: int, optional, default: 100
Maximum number of iterations or the number of epochs for SGD.
autoencoder_structure: list, default: None
The number of neurons of encoder/decoder layer for SDAE.
For example, autoencoder_structure = [200], the SDAE structure will be [vocab_size, 200, k, 200, vocab_size]
act_fn: str, default: 'relu'
Name of the activation function used for the auto-encoder.
Supported functions: ['sigmoid', 'tanh', 'elu', 'relu', 'relu6', 'leaky_relu', 'identity']
learning_rate: float, optional, default: 0.001
The learning rate for AdamOptimizer.
vocab_size: int, default: 8000
The size of text input for the SDAE.
lambda_u: float, optional, default: 0.1
The regularization parameter for users.
lambda_v: float, optional, default: 10
The regularization parameter for items.
lambda_w: float, optional, default: 0.1
The regularization parameter for SDAE weights.
lambda_n: float, optional, default: 1000
The regularization parameter for SDAE output.
a: float, optional, default: 1
The confidence of observed ratings.
b: float, optional, default: 0.01
The confidence of unseen ratings.
corruption_rate: float, optional, default: 0.3
The corruption ratio for input text of the SDAE.
dropout_rate: float, optional, default: 0.1
The probability that each element is removed in dropout of SDAE.
batch_size: int, optional, default: 128
The batch size for SGD.
trainable: boolean, optional, default: True
When False, the model is not trained and Cornac assumes that the model already
pre-trained (U and V are not None).
init_params: dictionary, optional, default: None
List of initial parameters, e.g., init_params = {'U':U, 'V':V}
U: ndarray, shape (n_users,k)
The user latent factors, optional initialization via init_params.
V: ndarray, shape (n_items,k)
The item latent factors, optional initialization via init_params.
seed: int, optional, default: None
Random seed for weight initialization.
References
----------
* Hao Wang, Naiyan Wang, Dit-Yan Yeung. CDL: Collaborative Deep Learning for Recommender Systems. In : SIGKDD. 2015. p. 1235-1244.
"""
def __init__(
self,
name="CDL",
k=50,
autoencoder_structure=None,
act_fn="relu",
lambda_u=0.1,
lambda_v=10,
lambda_w=0.1,
lambda_n=1000,
a=1,
b=0.01,
corruption_rate=0.3,
learning_rate=0.001,
vocab_size=8000,
dropout_rate=0.1,
batch_size=128,
max_iter=100,
trainable=True,
verbose=True,
init_params=None,
seed=None,
):
super().__init__(name=name, trainable=trainable, verbose=verbose)
self.k = k
self.lambda_u = lambda_u
self.lambda_v = lambda_v
self.lambda_w = lambda_w
self.lambda_n = lambda_n
self.a = a
self.b = b
self.corruption_rate = corruption_rate
self.dropout_rate = dropout_rate
self.learning_rate = learning_rate
self.vocab_size = vocab_size
self.name = name
self.max_iter = max_iter
self.autoencoder_structure = autoencoder_structure
self.act_fn = act_fn
self.batch_size = batch_size
self.verbose = verbose
self.seed = seed
self.rng = get_rng(seed)
# Init params if provided
self.init_params = {} if init_params is None else init_params
self.U = self.init_params.get("U", None)
self.V = self.init_params.get("V", None)
def _init(self):
n_users, n_items = self.num_users, self.num_items
if self.U is None:
self.U = xavier_uniform((n_users, self.k), self.rng)
if self.V is None:
self.V = xavier_uniform((n_items, self.k), self.rng)
[docs]
def fit(self, train_set, val_set=None):
"""Fit the model to observations.
Parameters
----------
train_set: :obj:`cornac.data.Dataset`, required
User-Item preference data as well as additional modalities.
val_set: :obj:`cornac.data.Dataset`, optional, default: None
User-Item preference data for model selection purposes (e.g., early stopping).
Returns
-------
self : object
"""
Recommender.fit(self, train_set, val_set)
self._init()
if self.trainable:
self._fit_cdl(train_set)
return self
def _fit_cdl(self, train_set):
import tensorflow.compat.v1 as tf
from .cdl import Model
tf.disable_eager_execution()
R = train_set.csc_matrix # csc for efficient slicing over items
text_feature = train_set.item_text.batch_bow(
np.arange(self.num_items)
) # bag-of-words features
text_feature = (text_feature - text_feature.min()) / (
text_feature.max() - text_feature.min()
) # normalization
# Build model
layer_sizes = (
[self.vocab_size]
+ self.autoencoder_structure
+ [self.k]
+ self.autoencoder_structure
+ [self.vocab_size]
)
tf.set_random_seed(self.seed)
model = Model(
n_users=self.num_users,
n_items=self.num_items,
n_vocab=self.vocab_size,
k=self.k,
layers=layer_sizes,
lambda_u=self.lambda_u,
lambda_v=self.lambda_v,
lambda_w=self.lambda_w,
lambda_n=self.lambda_n,
lr=self.learning_rate,
dropout_rate=self.dropout_rate,
U=self.U,
V=self.V,
act_fn=self.act_fn,
seed=self.seed,
)
# Training model
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
with tf.Session(config=config) as sess:
sess.run(tf.global_variables_initializer())
loop = trange(self.max_iter, disable=not self.verbose)
for _ in loop:
corruption_mask = self.rng.binomial(
1, 1 - self.corruption_rate, size=(self.num_items, self.vocab_size)
)
sum_loss = 0
count = 0
for i, batch_ids in enumerate(
train_set.item_iter(self.batch_size, shuffle=True)
):
batch_R = R[:, batch_ids]
batch_C = np.ones(batch_R.shape) * self.b
batch_C[batch_R.nonzero()] = self.a
feed_dict = {
model.text_mask: corruption_mask[batch_ids, :],
model.text_input: text_feature[batch_ids],
model.ratings: batch_R.A,
model.C: batch_C,
model.item_ids: batch_ids,
}
sess.run(model.opt1, feed_dict) # train U, V
_, _loss = sess.run(
[model.opt2, model.loss], feed_dict
) # train SDAE
sum_loss += _loss
count += len(batch_ids)
if i % 10 == 0:
loop.set_postfix(loss=(sum_loss / count))
self.U, self.V = sess.run([model.U, model.V])
tf.reset_default_graph()
if self.verbose:
print("Learning completed!")
[docs]
def score(self, user_idx, item_idx=None):
"""Predict the scores/ratings of a user for an item.
Parameters
----------
user_idx: int, required
The index of the user for whom to perform score prediction.
item_idx: int, optional, default: None
The index of the item for which to perform score prediction.
If None, scores for all known items will be returned.
Returns
-------
res : A scalar or a Numpy array
Relative scores that the user gives to the item or to all known items
"""
if self.is_unknown_user(user_idx):
raise ScoreException("Can't make score prediction for user %d" % user_idx)
if item_idx is not None and self.is_unknown_item(item_idx):
raise ScoreException("Can't make score prediction for item %d" % item_idx)
if item_idx is None:
return self.V.dot(self.U[user_idx, :])
return self.V[item_idx, :].dot(self.U[user_idx, :])
[docs]
def get_vector_measure(self):
"""Getting a valid choice of vector measurement in ANNMixin._measures.
Returns
-------
measure: MEASURE_DOT
Dot product aka. inner product
"""
return MEASURE_DOT
[docs]
def get_user_vectors(self):
"""Getting a matrix of user vectors serving as query for ANN search.
Returns
-------
out: numpy.array
Matrix of user vectors for all users available in the model.
"""
return self.U
[docs]
def get_item_vectors(self):
"""Getting a matrix of item vectors used for building the index for ANN search.
Returns
-------
out: numpy.array
Matrix of item vectors for all items available in the model.
"""
return self.V
