def __init__(self,
train_file,
test_file=None,
metadata_file=None,
similarity_matrix_file=None,
ranking_file=None,
neighbors=30,
rank_number=10,
similarity_metric="correlation",
space_type='\t'):
ItemKNN.__init__(self,
train_file,
test_file=test_file,
ranking_file=ranking_file,
neighbors=neighbors,
rank_number=rank_number,
similarity_metric=similarity_metric,
space_type=space_type)
if metadata_file is None and similarity_matrix_file is None:
print(
"This algorithm needs a similarity matrix or a metadata file!")
sys.exit(0)
if metadata_file is not None:
self.metadata = ReadFile(
metadata_file, space_type=space_type).read_metadata(self.items)
self.matrix = self.metadata['matrix'].T
self.similarity_matrix_file = similarity_matrix_file
def get_itemknn(self):
ItemKNN(train_file=self.train_file,
test_file=self.test_file,
output_file=self.output_file,
rank_length=self.rank_length,
similarity_metric=self.similarity_metric,
k_neighbors=int(self.k_neighbors),
sep=self.sep,
output_sep=self.output_sep).compute(verbose_evaluation=False)
def generate_recommendation(self):
self.ranking = []
for n, train_file in enumerate(self.gb_train_files):
if self.recommender == 'UserKNN':
rec = UserKNN(train_file=train_file,
similarity_metric=self.similarity_metric,
as_binary=True,
as_similar_first=False)
rec.compute(verbose=False, verbose_evaluation=False)
self.ranking += rec.ranking
elif self.recommender == 'ItemKNN':
rec = ItemKNN(train_file=train_file,
test_file=self.test_file,
similarity_metric=self.similarity_metric,
as_binary=True)
rec.compute(verbose=False, verbose_evaluation=False)
self.ranking += rec.ranking
elif self.recommender == 'MostPopular':
rec = MostPopular(train_file=train_file,
test_file=self.test_file,
as_binary=True)
rec.compute(verbose=False, verbose_evaluation=False)
self.ranking += rec.ranking
elif self.recommender == 'BPRMF':
rec = BprMF(train_file=train_file,
test_file=self.test_file,
batch_size=4)
rec.compute(verbose=False, verbose_evaluation=False)
self.ranking += rec.ranking
else:
raise ValueError(
'Error: Recommender not implemented or not exist!')
self.ranking = sorted(self.ranking, key=lambda x: (x[0], -x[2]))
from caserec.recommenders.item_recommendation.itemknn import ItemKNN
from caserec.recommenders.item_recommendation.most_popular import MostPopular
from caserec.utils.cross_validation import CrossValidation
db = 'u.data'
folds_path = ''
metrics = ('PREC', 'RECALL', 'NDCG', 'MAP')
recommender = ItemKNN()
CrossValidation(input_file=db,
recommender=recommender,
dir_folds=folds_path,
header=1,
k_folds=5).compute(metrics=metrics)
recommender = MostPopular()
CrossValidation(input_file=db,
recommender=recommender,
dir_folds=folds_path,
header=1,
k_folds=5).compute(metrics=metrics)
te_film = '../../../datasets/filmtrust/folds1/0/test.dat' metadata_item = '../../../datasets/ml-100k/metadata_movies.csv' sm_item = '../../../datasets/ml-100k/ItemSimilarity.csv' metadata_user = '******' sm_user = '******' label_set1 = '../../../datasets/filmtrust/folds1/0/labeled_set_1.dat' label_set2 = '../../../datasets/filmtrust/folds1/0/labeled_set_2.dat' ensemble = '../../../datasets/filmtrust/folds1/0/ensemble_set.dat' train_list = [tr] '''单独运行''' UserKNN(tr_film, te_film, as_similar_first=True).compute() ItemKNN(tr_film, te_film, as_similar_first=True).compute() # ContentBased(tr, te, similarity_file=sm_item).compute() # BprMF(tr, te).compute() # MostPopular(tr, te, as_binary=True).compute() # PaCoRecommender(tr, te, as_binary=True).compute() # ItemAttributeKNN(tr, te, similarity_file=sm_item, as_similar_first=True).compute() # UserAttributeKNN(tr, te, similarity_file=sm_user, as_similar_first=True).compute() # GroupBasedRecommender(train_list, te).compute() '''交叉验证'''
"""
# # Cross Validation
# recommender = UserKNN(as_binary=True)
#
# CrossValidation(input_file=db, recommender=recommender, dir_folds=folds_path, header=1, k_folds=5).compute()
#
# # Simple
# UserKNN(tr, te, as_binary=True).compute()
# UserAttributeKNN(tr, te, metadata_file=metadata_user).compute()
# UserAttributeKNN(tr, te, similarity_file=sm_user).compute()
"""
ItemKNN
"""
# # Cross Validation
recommender = ItemKNN(as_binary=True)
#
CrossValidation(input_file=db,
recommender=recommender,
dir_folds=folds_path,
header=1,
k_folds=5).compute()
#
# # Simple
# ItemKNN(tr, te, as_binary=True).compute()
ItemAttributeKNN(tr, te, metadata_file=metadata_item).compute()
# ItemAttributeKNN(tr, te, similarity_file=sm_item).compute()