# cnn2vec = loadData_DKRL("./data/1000nepoches/res_3_4_1000/cnn_vec_out.txt", "./data/train_data/entity2id.txt")
# rel2vec = loadData_DKRL("./data/1000nepoches/res_3_4_1000/relation2vec.bern", "./data/train_data/relation2id.txt")
# stru2vec = loadData_DKRL("./data/1000nepoches/res_3_4_1000/entity2vec.bern", "./data/train_data/entity2id.txt")
# test_num, test = openTxt.openTrain("./data/train_data/test_3_4.txt")
# train_num, train = openTxt.openTrain("./data/train_data/train_3_4.txt")
cnn2vec = loadData_DKRL(
"./data/EAnci/1000nepoches/res_4_7_1000/cnn_vec_out.txt",
"./data/train_data/entity2id.txt")
rel2vec = loadData_DKRL(
"./data/EAnci/1000nepoches/res_4_7_1000/relation2vec.bern",
"./data/train_data/relation2id.txt")
stru2vec = loadData_DKRL(
"./data/EAnci/1000nepoches/res_4_7_1000/entity2vec.bern",
"./data/train_data/entity2id.txt")
test_num, test = openTxt.openTrain("./data/train_data/test_4_7.txt")
train_num, train = openTxt.openTrain("./data/train_data/train_4_7.txt")
###new_y_predict = SameInheritClass_Recovery.SameInheritClassRecovery(x_test_triple,y_predict,train)
###precissio, recall, acc, f1 = Result.get_result(y_test, new_y_predict)
###三元组分类
# print("rate = {}".format(0.85))
# y_predict_cnn,y_predict_multi = TripleClassfication.TripleClassfication(train,x_test_triple,y_predict,cnn2vec,stru2vec,rel2vec,0.85)
# precissio, recall, acc, f1 = Result.get_result(y_test, y_predict_cnn)
# precissio, recall, acc, f1 = Result.get_result(y_test, y_predict_multi)
sum1, sum2, sum3, sum4 = 0, 0, 0, 0
sum5, sum6, sum7, sum8 = 0, 0, 0, 0
for i in range(0, 10):
x_train, y_train, x_test, y_test, x_test_triple = generateClassiferData(
train, test, cnn2vec, stru2vec, 1, (4 / 7))
y_predict = Phase3Classifier.train(x_train, y_train, x_test, y_test, 1)
discriminant_analysis, random_projection)
if __name__ == '__main__':
cnn2vec = eTour_Experiment.loadData_DKRL(
"./data/eTour/dkrl_result/res_eTour_1200_100_H_M_Improve/cnn_vec_out.txt",
"./data/eTour/train_data/entity2id.txt")
rel2vec = eTour_Experiment.loadData_DKRL(
"./data/eTour/dkrl_result/res_eTour_1200_100_H_M_Improve/relation2vec.bern",
"./data/eTour/train_data/relation2id.txt")
stru2vec = eTour_Experiment.loadData_DKRL(
"./data/eTour/dkrl_result/res_eTour_1200_100_H_M_Improve/entity2vec.bern",
"./data/eTour/train_data/entity2id.txt")
# test_num, test = openTxt.openTrain("./data/eTour/train_data/testZ.txt")
# train_num, train = openTxt.openTrain("./data/eTour/train_data/trainZ.txt")
# oracle_num,oracleList = openTxt.openOracle("./data/eTour/oracle/oracle line.txt")
dkrl_train_num, dkrl_train = openTxt.openTrain(
"./data/eTour/train_data/DKRL_trainZ.txt")
dkrl_test_num, dkrl_test = openTxt.openTrain(
"./data/eTour/train_data/DKRL_testZ.txt")
trainZ_N_num, trainZ_N = openTxt.openTrain(
"./data/eTour/train_data/trainZ_N.txt")
ClassiferTestTriples_Num, ClassiferTestTriples = openTxt.openTrain(
"./data/eTour/train_data/Classifer_test_Z.txt")
entity_vec = []
entity_type = []
for triple in dkrl_train + trainZ_N:
if triple[2] == "oracle_link":
if stru2vec[triple[0]] not in entity_vec:
entity_vec.append(stru2vec[triple[0]])
entity_type.append(1)
if stru2vec[triple[1]] not in entity_vec:
entity_vec.append(stru2vec[triple[1]])
if w in bigWord:
newText += ("_" + w)
else:
newText += w
temp_list = newText.split("_")
word_list = []
for word in temp_list:
if len(word) == 0:
continue
else:
word_list.append(word)
return word_list
if __name__ == '__main__':
xmlParseTripleNum, xmlParseTriple = openTxt.openTrain(
"./data/eTour/eTour_xmlParseResult.txt")
code_entity_list = []
word_dictionary = []
for triple in xmlParseTriple:
if triple[0] not in code_entity_list:
code_entity_list.append(triple[0])
if triple[1] not in code_entity_list:
code_entity_list.append(triple[1])
code_entity_word_dict = {}
for entity in code_entity_list:
code_entity_word_dict[entity] = NameSplit(entity)
for word in code_entity_word_dict[entity]:
if word not in word_dictionary:
word_dictionary.append(word)
word_num, word_i2e_dict = openTxt.openEntity2Des(
"./data/convert_word_list.txt")
import nltk
import Tool
import openTxt
import random
if __name__ == '__main__':
file_num, eTourCC2class = openTxt.eTourCC2class_open(
"./data/eTour/eTour_ea2class.txt")
word_num, codeEntityWords = openTxt.openTrain(
"./data/eTour/prosessData/codeEntityWords.txt")
word_num, codeEntity2Words = openTxt.openEntity2Des(
"./data/eTour/prosessData/codeEntityWords.txt")
xmlParser_num, xmlParser = openTxt.openTrain(
"./data/eTour/eTour_xmlParseResult.txt")
oracle_num, temp_oracleList = openTxt.openOracle(
"./data/eTour/oracle/oracle line.txt")
ucLineNum, uc2ucLine = openTxt.openTrain(
"./data/eTour/prosessData/uc2ucLink.txt")
uc_num, ucEntityWords = openTxt.openTrain(
"./data/eTour/prosessData/ucEntityWords.txt")
oracleList = []
# ucName2file = {}
# for file in eTourCC2class:
# ucName2file[eTourCC2class[file]["class"][0]] = file
# print(ucName2file["Beniculturali"])
dirList = ["One", "Two", "Three"]
crossValidationNum = 3
for triple in temp_oracleList:
oracleList.append(
(triple[0], eTourCC2class[triple[1]]["class"][0], triple[2]))
import openTxt
if __name__ == '__main__':
triple_num,triple_List = openTxt.openTrain("./data/entityWords.txt")
word_list = []
for triple in triple_List:
for word in triple[2].strip().split(" "):
if word not in word_list:
word_list.append(word)
with open("./data/word2id.txt","w",encoding="utf-8") as file:
for index,word in enumerate(word_list):
file.write(word+"\t"+str(index)+"\n")
import nltk
import NameSplit
import openTxt
import Tool
import re
if __name__ == '__main__':
class_num,class2comment = openTxt.openDescription("./data/eTour/class2comment.txt")
file_num,eTourCC2class = openTxt.eTourCC2class_open("./data/eTour/eTour_ea2class.txt")
entity_num,entity2word = openTxt.openEntity2Des("./data/eTour/code_entity_word_dict.txt")
word_num,temp_word_i2e_dict = openTxt.openEntity2Des("./data/convert_word_list.txt")
uc_num,ucEntityWords = openTxt.openTrain("./data/eTour/prosessData/uc2des.txt")
word_i2e_dict = {}
saveEnglish = re.compile("[^a-zA-Z]")
classList = [tu[0] for tu in class2comment]
for key in entity2word:
if key in classList:
index = classList.index(key)
class2comment[index][1] += (" "+key)
else:
class2comment.append((key,key))
for key in temp_word_i2e_dict.keys():
if key not in word_i2e_dict.keys():
word_i2e_dict[key] = temp_word_i2e_dict[key]
if key.lower() not in word_i2e_dict.keys():
word_i2e_dict[key.lower()] = temp_word_i2e_dict[key]
if key.capitalize() not in word_i2e_dict.keys():
word_i2e_dict[key.capitalize()] = temp_word_i2e_dict[key]