def __init__(self, concepts, settings):
from SimilarityCalculator import SimilarityCalculator
self.sc = SimilarityCalculator(settings)
from WordNetSynonyms import WordNetSynonyms
self.wns = WordNetSynonyms()
self.totalThreshold = settings.SimilarityThreshold
def TestPipeline():
from SimilarityCalculator import SimilarityCalculator
sc = SimilarityCalculator(settings)
from WordNetSynonyms import WordNetSynonyms
wns = WordNetSynonyms()
while True:
wordA = GetInputWord(1)
if wordA != "n":
wordB = GetInputWord(2)
if wordB != "n":
if settings.Reload is True:
sc.ReloadSettings(settings)
similarity = sc.GetSimilarity(wordA, wordB)
print("Similarity: %f" % similarity)
if similarity > settings.SimilarityThreshold:
synonymity = wns.IsSynonym(wordA, wordB)
print("Synonymity: %s" % synonymity)
if synonymity is True:
print("The words are both accepted in the model!")
else:
print(
"One of the words is not accepted. They are synonyms."
)
else:
print("The words are both accepted in the model!")
else:
break
else:
break
class SynonymRemover(object):
"""Removes synonyms using a variety of methods"""
def __init__(self, concepts, settings):
from SimilarityCalculator import SimilarityCalculator
self.sc = SimilarityCalculator(settings)
from WordNetSynonyms import WordNetSynonyms
self.wns = WordNetSynonyms()
self.totalThreshold = settings.SimilarityThreshold
# Returns all items in newCollection that don't have a synonym in oldCollection.
def RemoveSynonyms(self, oldCollection, newCollection):
noSynonyms = []
for newItem in newCollection:
print("Comparing %s to collection:" % newItem) # DEBUG
if (not self.IsSynonym(newItem, oldCollection)):
noSynonyms.append(newItem)
return noSynonyms
# Returns whether a synonym of item is in collection.
def IsSynonym(self, item, collection):
similarity = self.sc.GetMaxSimilarity(item, collection)
if similarity >= self.totalThreshold:
if self.wns.HasSynonym(item, collection):
return True
else:
return False
else:
return False
def calculateSimilarity(self):
similarityCalculator = SimilarityCalculator(self.__movieUserMatrix)
self.__similarityMatrix = similarityCalculator.calculateSimilarity()
def __init__(self, settings, loadAll=False):
self.s = settings
from SimilarityCalculator import SimilarityCalculator
self.sc = SimilarityCalculator(self.s, loadAll=loadAll)
class CompoundHandler(object):
"""Handles compound terms."""
import Settings
def __init__(self, settings, loadAll=False):
self.s = settings
from SimilarityCalculator import SimilarityCalculator
self.sc = SimilarityCalculator(self.s, loadAll=loadAll)
def GetSimilarity(self, compoundA, compoundB):
wordsA = compoundA.split()
wordsB = compoundB.split()
if len(wordsA) > 2 or len(wordsB) > 2:
print("We don't support compounds larger than two words yet.")
return -1.0
if len(wordsA) == 0 or len(wordsB) == 0:
print("One of the compounds is empty.")
return -1.0
compoundAisCompound = len(wordsA) > 1
compoundBisCompound = len(wordsB) > 1
# Level - Altitude
if compoundAisCompound is not True and compoundBisCompound is not True:
# print("Comparing %s to %s:" % (wordsA[0], wordsB[0])) # DEBUG
similarity = self.sc.GetSimilarity(wordsA[0], wordsB[0])
print("Similarity: %s <> %s = %f" %
(wordsA[0], wordsB[0], similarity)) # DEBUG
# Flight Level - Altitude
elif compoundAisCompound is True and compoundBisCompound is not True:
# print("Comparing %s to %s:" % (wordsA[0], wordsB[0])) # DEBUG
similarityA = self.sc.GetSimilarity(wordsA[0],
wordsB[0]) # Flight - Altitude
print("similarityA: %s <> %s = %f" %
(wordsA[0], wordsB[0], similarityA)) # DEBUG
# print("Comparing %s to %s:" % (wordsA[1], wordsB[0])) # DEBUG
similarityB = self.sc.GetSimilarity(wordsA[1],
wordsB[0]) # Level - Altitude
print("similarityB: %s <> %s = %f" %
(wordsA[1], wordsB[0], similarityB)) # DEBUG
similarity = (self.s.Alpha * similarityA +
self.s.Beta * similarityB)
print("Similarity: %f * %f + %f * %f = %f" %
(self.s.Alpha, similarityA, self.s.Beta, similarityB,
similarity)) # DEBUG
# Level - Flight Altitude
elif compoundAisCompound is not True and compoundBisCompound is True:
# print("Comparing %s to %s:" % (wordsA[0], wordsB[0])) # DEBUG
similarityA = self.sc.GetSimilarity(wordsA[0],
wordsB[0]) # Level - Flight
print("similarityA: %s <> %s = %f" %
(wordsA[0], wordsB[0], similarityA)) # DEBUG
# print("Comparing %s to %s:" % (wordsA[0], wordsB[1])) # DEBUG
similarityB = self.sc.GetSimilarity(wordsA[0],
wordsB[1]) # Level - Altitude
print("similarityB: %s <> %s = %f" %
(wordsA[0], wordsB[1], similarityB)) # DEBUG
similarity = (self.s.Alpha * similarityA +
self.s.Beta * similarityB)
print("Similarity: %f * %f + %f * %f = %f" %
(self.s.Alpha, similarityA, self.s.Alpha, similarityB,
similarity)) # DEBUG
# Flight Level - Flight Altitude
else:
# print("Comparing %s to %s:" % (wordsA[0], wordsB[0])) # DEBUG
similarityC = self.sc.GetSimilarity(wordsA[0],
wordsB[0]) # Flight - Flight
print("similarityC: %s <> %s = %f" %
(wordsA[0], wordsB[0], similarityC)) # DEBUG
# print("Comparing %s to %s:" % (wordsA[1], wordsB[1])) # DEBUG
similarityD = self.sc.GetSimilarity(wordsA[1],
wordsB[1]) # Level - Altitude
print("similarityD: %s <> %s = %f" %
(wordsA[1], wordsB[1], similarityD)) # DEBUG
# print("Comparing %s to %s:" % (wordsA[0], wordsB[1])) # DEBUG
similarityEa = self.sc.GetSimilarity(
wordsA[0], wordsB[1]) # Flight - Altitude
print("similarityEa: %s <> %s = %f" %
(wordsA[0], wordsB[1], similarityEa)) # DEBUG
# print("Comparing %s to %s:" % (wordsA[1], wordsB[0])) # DEBUG
similarityEb = self.sc.GetSimilarity(wordsA[1],
wordsB[0]) # Level - Flight
print("similarityEb: %s <> %s = %f" %
(wordsA[1], wordsB[0], similarityEb)) # DEBUG
similarity = (self.s.Gamma * similarityC +
self.s.Delta * similarityD +
self.s.Epsilon * similarityEa +
self.s.Epsilon * similarityEb)
print(
"Similarity: (%f * %f) + (%f * %f) + (%f * %f) + (%f * %f) = %f"
% (self.s.Gamma, similarityC, self.s.Delta, similarityD,
self.s.Epsilon, similarityEa, self.s.Epsilon, similarityEb,
similarity)) # DEBUG
return similarity
def ReloadSettings(self, settings):
self.s = settings
self.sc.ReloadSettings(settings)
def calculateSimilarity(self): similarityCalculator = SimilarityCalculator(self.__userMovieMatrix) self.__similarityMatrix = similarityCalculator.calculateSimilarity()
from SimilarityCalculator import SimilarityCalculator s = SimilarityCalculator(2)