class LMModel:
def __init__(self, seqExtractor, n = 2):
self.words = None
self.seqExtractor = seqExtractor
self.lm0 = KneserNeyLM(n)
self.lm1 = KneserNeyLM(n)
def train(self, rows):
self.seqExtractor.train(rows)
self.lm0.train([self.seqExtractor.extract(row) for row in rows if not row.insult])
self.lm1.train([self.seqExtractor.extract(row) for row in rows if row.insult])
def classify1(self, row):
seq = self.seqExtractor.extract(row)
w = 0.0 + self.lm1.score(seq) - self.lm0.score(seq)
if 100 < w: w = 100
if w < -100: w = -100
return 1.0/(1.0 + math.exp(-w))
def classify(self, rows):
return array([self.classify1(row) for row in rows])
def __init__(self, seqExtractor, n = 2): self.words = None self.seqExtractor = seqExtractor self.lm0 = KneserNeyLM(n) self.lm1 = KneserNeyLM(n)
