from __future__ import division
import re
from numpy import ones, array
from numpy.lib.scimath import log
from nltk import *
def loadDataSet():
pos=open("pos_train.txt",'r')
neg=open("neg_train.txt",'r')
lst_all=[]
classVec=[]
for i in range(700):
classVec.append(i%2)
for i in range(350):
str0=pos.readline()
str1=neg.readline()
regEx0=re.compile('\\W*')
regEx1=re.compile('\\W*')
lst_pos=regEx0.split(str0)
lst_neg=regEx1.split(str1)
lst_all.append([tok.lower() for tok in lst_pos if len(tok)>0])
lst_all.append([tok.lower() for tok in lst_neg if len(tok)>0])
return lst_all,classVec
def loadTestSet():
pos=open("pos_test.txt",'r')
neg=open("neg_test.txt",'r')
lst_pos_test=[]
lst_neg_test=[]
for i in range(350):
str0=pos.readline()
regEx0=re.compile('\\W*')
lst_pos=regEx0.split(str0)
lst_pos_test.app
Python贝叶斯算法进行情感分析
最新推荐文章于 2025-05-10 13:01:19 发布