def BehaviorBasedCF():
np.random.seed(0)
random.seed(0)
# Load up common data set for the recommender algorithms
(ml, evaluationData, rankings) = MyDump.LoadMovieLensData(loader)
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(evaluationData, rankings, load=True)
# User-based KNN
UserKNN = KNNBasic(sim_options={'name': 'cosine', 'user_based': True})
evaluator.AddAlgorithm(UserKNN, "User KNN")
# Item-based KNN
ItemKNN = KNNBasic(sim_options={'name': 'cosine', 'user_based': False})
evaluator.AddAlgorithm(ItemKNN, "Item KNN")
# Just make random recommendations
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
evaluator.Evaluate(
False
) # load is also False, cause simsMatrix needs to be loaded; I haven't figured it out.
evaluator.SampleTopNRecs(ml)
def ContentRecs():
""" for this content-based(item) recommender
calculate items' cosine similarity matrix in alg.fit()
I don't test `HitRate` for `topN` recommendation, because here it's impossible to do that. What I recommend by this algorithm are all the movies the user haven't rated.
"""
np.random.seed(0)
random.seed(0)
# loader = True
# Load up common data set for the recommender algorithms
# print(f'call ContentRecs()\nloader = {loader}')
(ml, evaluationData, rankings) = MyDump.LoadMovieLensData(loader)
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(evaluationData, rankings, load=True)
contentKNN = ContentKNNAlgorithm()
evaluator.AddAlgorithm(contentKNN, "ContentKNN")
# Just make random recommendations
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
evaluator.Evaluate(False, True) # not topN, able load
# recommend 10(default) items
evaluator.SampleTopNRecs(ml)
def BuildRecModel():
np.random.seed(0)
random.seed(0)
# Load up common data set for the recommender algorithms
(evaluationData, rankings) = LoadData()
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(evaluationData, rankings)
hotelKNNAlgorithm = HotelKNNAlgorithm()
log.info('Algorithm=K-Nearest Neighbour')
evaluator.SetAlgorithm(hotelKNNAlgorithm, 'HotelKNNAlgorithm')
evaluator.Evaluate(False)
evaluator.TrainAndSaveAlgorithm()
def AutoRec():
np.random.seed(0)
random.seed(0)
ml, evaluationData, rankings = MyDump.LoadMovieLensData(loader)
evaluator = Evaluator(evaluationData, rankings, loader)
myAutoRec = RBMAlgorithm()
evaluator.AddAlgorithm(myAutoRec, "AutoRec")
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
evaluator.Evaluate(doTopN=False, load=loader)
evaluator.SampleTopNRecs(ml, loader)
def CompareSVDRandom():
np.random.seed(0)
random.seed(0)
# loader = True
start_t = time.time()
# Load up common data set for the recommender algorithms
# (evaluationData, rankings) = LoadMovieLensData()
# MyDump.Save('ratingsDataset', data = evaluationData, verbose = 1)
# MyDump.Save('rankings', data = rankings, verbose = 1)
# _,_,evaluationData = MyDump.Load('ratingsDataset', 1)
# _,_,rankings = MyDump.Load('rankings',1)
# if evaluationData == None or rankings == None:
# (_, evaluationData, rankings) = MyDump.LoadMovieLensData() # meat
# MyDump.Save('ratingsDataset', data = evaluationData, verbose = 1)
# MyDump.Save('rankings', data = rankings, verbose = 1)
_, evaluationData, rankings = MyDump.LoadMovieLensData(loader)
print(
f'------time consumption: {time.time() - start_t} for LoadMovieLensData()'
)
start_t = time.time()
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(dataset=evaluationData,
rankings=rankings,
load=loader)
print(
f'------time consumption: {time.time() - start_t} for create an evaluator instance'
)
start_t = time.time()
# Throw in an SVD recommender
SVDAlgorithm = SVD(random_state=10)
evaluator.AddAlgorithm(SVDAlgorithm, "SVD")
# Just make random recommendations
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
start_t = time.time()
evaluator.Evaluate(True, loader) # doTopN, loader
print(
f'------time consumption: {time.time() - start_t} for evaluator.Evaluate()'
)
start_t = time.time()
def MF():
""" the idea behind is math, latent features
implementation is simple out of library source
"""
np.random.seed(0)
random.seed(0)
ml, evaluationData, rankings = MyDump.LoadMovieLensData(loader)
evaluator = Evaluator(evaluationData, rankings, loader)
mySVD = SVD(random_state=10)
evaluator.AddAlgorithm(mySVD, "SVD") # the same with before
mySVDpp = SVDpp(random_state=10)
evaluator.AddAlgorithm(mySVDpp, "SVDpp")
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
evaluator.Evaluate(doTopN=False, load=loader)
evaluator.SampleTopNRecs(ml, loader)
def contentf(self):
# Load up common data set for the recommender algorithms
(ml, evaluationData, rankings) = self.LoadNovieLensData()
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(evaluationData, rankings)
contentKNN = ContentKNNAlgorithm()
evaluator.AddAlgorithm(contentKNN, "ContentKNN")
# Just make random recommendations
#Random = NormalPredictor()
#evaluator.AddAlgorithm(Random, "Random")
evaluator.Evaluate(False)
evaluator.SampleTopNRecs(ml)
def RBMtest():
np.random.seed(0)
random.seed(0)
ml, evaluationData, rankings = MyDump.LoadMovieLensData(loader)
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(evaluationData, rankings, loader)
# RBM
# able to tune by trying more parameter combination
myRBM = RBMAlgorithm(epochs=20)
evaluator.AddAlgorithm(myRBM, "RBM")
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
evaluator.Evaluate(doTopN=False, load=loader)
evaluator.SampleTopNRecs(ml, loader)
def contentGive(id):
np.random.seed(0)
random.seed(0)
# Load up common data set for the recommender algorithms
(ml, evaluationData, rankings) = LoadMovieLensData()
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(evaluationData, rankings)
contentKNN = ContentKNNAlgorithm()
evaluator.AddAlgorithm(contentKNN, "ContentKNN")
# Just make random recommendations
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
evaluator.Evaluate(False)
evaluator.SampleTopNRecs(ml, id)
)
rankings = ml.getPopularityRanks()
return (ml, data, rankings)
np.random.seed(0)
random.seed(0)
# Load up common data set for the recommender algorithms
(ml, evaluationData, rankings) = LoadMovieLensData()
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(evaluationData, rankings)
# SVD
SVD = SVD()
evaluator.AddAlgorithm(SVD, "SVD")
# SVD++
SVDPlusPlus = SVDpp()
evaluator.AddAlgorithm(SVDPlusPlus, "SVD++")
# Just make random recommendations
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
# Fight!
evaluator.Evaluate(False)
evaluator.SampleTopNRecs(ml)
print("Loading movie ratings...")
data = ml.loadMovieLensLatestSmall()
print(
"\nComputing movie popularity ranks so we can measure novelty later..."
)
rankings = ml.getPopularityRanks()
return (ml, data, rankings)
np.random.seed(0)
random.seed(0)
# Load up common data set for the recommender algorithms
(ml, evaluationData, rankings) = LoadMovieLensData()
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(evaluationData, rankings)
#Autoencoder
AutoRec = AutoRecAlgorithm()
evaluator.AddAlgorithm(AutoRec, "AutoRec")
# Just make random recommendations
Random = NormalPredictor()
evaluator.AddAlgorithm(Random, "Random")
# Fight!
evaluator.Evaluate(True)
evaluator.SampleTopNRecs(ml)
print("Loading movie ratings...")
data = source.loadMovieLensRating()
print("Prepare movie information...")
source.computeMovieInformation()
print("Creating ranking for each movie ...")
rankings = source.getPopularityRanksByRating()
return (source, data, rankings)
np.random.seed(0)
random.seed(0)
# Load up common data set for the recommender algorithms
(dataSource, data, rankings) = LoadData()
# Construct an Evaluator to, you know, evaluate them
evaluator = Evaluator(data, rankings)
contentKNN = KNNAlgorithm()
evaluator.AddAlgorithm(contentKNN, "ContentKNN")
# Just make random recommendations
# Random = NormalPredictor()
# evaluator.AddAlgorithm(Random, "Random")
evaluator.Evaluate()
useTargetId = 85
totalMovieNeeded = 5
evaluator.GetRecomendationMovie(dataSource, useTargetId, totalMovieNeeded)
