def __init__(self, sess, dataset, conf):
super(DeepICF, self).__init__(dataset, conf)
self.pretrain_file = conf["pretrain_file"]
self.verbose = conf["verbose"]
self.batch_size = conf["batch_size"]
self.use_batch_norm = conf["batch_norm"]
self.num_epochs = conf["epochs"]
self.weight_size = conf["weight_size"]
self.embedding_size = conf["embedding_size"]
self.n_hidden = conf["layers"]
regs = conf["regs"]
self.reg_W = conf["regw"]
self.lambda_bilinear = regs[0]
self.gamma_bilinear = regs[1]
self.eta_bilinear = regs[2]
self.alpha = conf["alpha"]
self.beta = conf["beta"]
self.num_negatives = conf["num_neg"]
self.learning_rate = conf["learning_rate"]
self.activation = conf["activation"]
self.algorithm = conf["algorithm"]
self.learner = conf["learner"]
self.embed_init_method = conf["embed_init_method"]
self.weight_init_method = conf["weight_init_method"]
self.bias_init_method = conf["bias_init_method"]
self.stddev = conf["stddev"]
self.dataset = dataset
self.num_items = dataset.num_items
self.num_users = dataset.num_users
self.train_dict = csr_to_user_dict(dataset.train_matrix)
self.sess = sess
def get_user_train_dict(self, by_time=False):
if by_time:
train_dict = csr_to_user_dict_bytime(self.time_matrix, self.train_matrix)
else:
train_dict = csr_to_user_dict(self.train_matrix)
return train_dict
def __init__(self, sess, dataset, conf):
super(NAIS, self).__init__(dataset, conf)
self.pretrain = conf["pretrain"]
self.verbose = conf["verbose"]
self.batch_size = conf["batch_size"]
self.num_epochs = conf["epochs"]
self.weight_size = conf["weight_size"]
self.embedding_size = conf["embedding_size"]
self.data_alpha = conf["data_alpha"]
self.regs = conf["regs"]
self.is_pairwise = conf["is_pairwise"]
self.topK = conf["topk"]
self.lambda_bilinear = self.regs[0]
self.gamma_bilinear = self.regs[1]
self.eta_bilinear = self.regs[2]
self.alpha = conf["alpha"]
self.beta = conf["beta"]
self.num_negatives = conf["num_neg"]
self.learning_rate = conf["learning_rate"]
self.activation = conf["activation"]
self.loss_function = conf["loss_function"]
self.algorithm = conf["algorithm"]
self.learner = conf["learner"]
self.embed_init_method = conf["embed_init_method"]
self.weight_init_method = conf["weight_init_method"]
self.stddev = conf["stddev"]
self.pretrain_file = conf["pretrain_file"]
self.dataset = dataset
self.num_items = dataset.num_items
self.num_users = dataset.num_users
self.train_dict = csr_to_user_dict(self.dataset.train_matrix)
self.sess = sess
def __init__(self, sess, dataset, conf):
super(DiffNet, self).__init__(dataset, conf)
self.learning_rate = conf["learning_rate"]
self.embedding_size = conf["embedding_size"]
self.learner = conf["learner"]
self.loss_function = conf["loss_function"]
self.num_epochs = conf["epochs"]
self.reg_mf = conf["reg_mf"]
self.batch_size = conf["batch_size"]
self.num_negatives = conf["num_negatives"]
self.user_feature_file = conf["user_feature_file"]
self.item_feature_file = conf["item_feature_file"]
self.feature_dimension = conf["feature_dimension"]
self.init_method = conf["init_method"]
self.stddev = conf["stddev"]
self.verbose = conf["verbose"]
self.num_users = dataset.num_users
self.num_items = dataset.num_items
self.userids = dataset.userids
self.itemids = dataset.itemids
self.dataset = dataset
self.trainMatrix = dataset.trainMatrix
self.trainDict = csr_to_user_dict(self.trainMatrix)
self.social_matrix = self.social_matrix + self.social_matrix.transpose()
self.consumed_items_sparse_matrix, self.social_neighbors_sparse_matrix = self.input_supply()
self.sess = sess
def __init__(self, sess, dataset, conf):
super(DAE, self).__init__(dataset, conf)
self.hidden_neuron = conf["hidden_neuron"]
self.learning_rate = conf["learning_rate"]
self.learner = conf["learner"]
self.reg = conf["reg"]
self.num_epochs = conf["epochs"]
self.batch_size = conf["batch_size"]
self.verbose = conf["verbose"]
self.h_act = conf["h_act"]
self.g_act = conf["g_act"]
self.corruption_level = conf["corruption_level"]
self.init_method = conf["init_method"]
self.stddev = conf["stddev"]
self.num_users = dataset.num_users
self.num_items = dataset.num_items
self.dataset = dataset
self.train_dict = csr_to_user_dict(dataset.train_matrix)
self.sess = sess
def __init__(self, sess, dataset, conf):
super(MultiDAE, self).__init__(dataset, conf)
self.learning_rate = conf["learning_rate"]
self.learner = conf["learner"]
self.batch_size = conf["batch_size"]
self.dataset = dataset
self.num_users = dataset.num_users
self.num_items = dataset.num_items
self.p_dims = conf["p_dim"] + [self.num_items]
self.q_dims = self.p_dims[::-1]
self.dims = self.q_dims + self.p_dims[1:]
self.act = conf["activation"]
self.reg = conf["reg"]
self.num_epochs = conf["epochs"]
self.weight_init_method = conf["weight_init_method"]
self.bias_init_method = conf["bias_init_method"]
self.stddev = conf["stddev"]
self.verbose = conf["verbose"]
self.train_dict = csr_to_user_dict(dataset.train_matrix)
self.sess = sess
def __init__(self, sess, dataset, conf):
super(SBPR, self).__init__(dataset, conf)
self.learning_rate = conf["learning_rate"]
self.embedding_size = conf["embedding_size"]
self.learner = conf["learner"]
self.loss_function = conf["loss_function"]
self.num_epochs = conf["num_epochs"]
self.reg_mf = conf["reg_mf"]
self.batch_size = conf["batch_size"]
self.init_method = conf["init_method"]
self.stddev = conf["stddev"]
self.verbose = conf["verbose"]
self.dataset = dataset
self.num_users = dataset.num_users
self.num_items = dataset.num_items
self.userids = self.dataset.userids
self.train_dict = csr_to_user_dict(self.dataset.train_matrix)
# self.social_matrix = self._get_social_matrix()
self.userSocialItemsSetList = self._get_SocialItemsSet()
self.sess = sess
def __init__(self, sess, dataset, conf):
super(FISM, self).__init__(dataset, conf)
self.batch_size = conf["batch_size"]
self.num_epochs = conf["epochs"]
self.embedding_size = conf["embedding_size"]
self.lambda_bilinear = conf["lambda"]
self.gamma_bilinear = conf["gamma"]
self.alpha = conf["alpha"]
self.num_negatives = conf["num_neg"]
self.learning_rate = conf["learning_rate"]
self.learner = conf["learner"]
self.topK = conf["topk"]
self.loss_function = conf["loss_function"]
self.is_pairwise = conf["is_pairwise"]
self.num_negatives = conf["num_neg"]
self.init_method = conf["init_method"]
self.stddev = conf["stddev"]
self.verbose = conf["verbose"]
self.num_users = dataset.num_users
self.num_items = dataset.num_items
self.dataset = dataset
self.train_dict = csr_to_user_dict(self.dataset.train_matrix)
self.sess = sess
def get_user_test_neg_dict(self):
test_neg_dict = None
if self.negative_matrix is not None:
test_neg_dict = csr_to_user_dict(self.negative_matrix)
return test_neg_dict
def get_user_test_dict(self):
test_dict = csr_to_user_dict(self.test_matrix)
return test_dict
tf.set_random_seed(2017)
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
if __name__ == "__main__":
is_windows = sys.platform.startswith('win')
if is_windows:
root_folder = 'XXXXXX/PythonProjects/SGL/'
else:
root_folder = 'XXXXXX/PythonProjects/SGL/'
conf = Configurator(root_folder + "NeuRec.properties",
default_section="hyperparameters")
dataset = Dataset(conf)
num_users = dataset.num_users
num_items = dataset.num_items
train_dict = tool.csr_to_user_dict(dataset.train_matrix)
test_dict = tool.csr_to_user_dict(dataset.test_matrix)
num_trainings = dataset.train_matrix.nnz
count = 0
while count < num_trainings * conf.ratio:
u_id = np.random.randint(num_users)
i_id = np.random.randint(num_items)
if i_id not in train_dict[u_id]:
if u_id not in test_dict:
train_dict[u_id].append(i_id)
count += 1
else:
if i_id not in test_dict[u_id]:
train_dict[u_id].append(i_id)
count += 1
with open(
def get_user_val_neg_dict(self):
val_neg_dict = None
if self.negative_matrix is not None:
val_neg_dict = csr_to_user_dict(self.negative_matrix)
return val_neg_dict
