Ejemplo n.º 1
0
def xahoi_thethao_mix():
	print("xahoi_thethao_mix")
	train_files = ["../data_conll_topic/News_Train/Xa_hoi.muc",
		"../data_conll_topic/News_Train/The_thao.muc"]
	dev_files = ["../data_conll_topic/News_Dev/The_thao.muc"]
	test_files = ["../data_conll_topic/News_Test/The_thao.muc"]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 2
0
def kinhte_thethao_mix_no_shuffle():
	print("kinhte_thethao_mix_no_shuffle")
	train_files = ["../data_conll_topic/News_Train/Kinh_te.muc",
		"../data_conll_topic/News_Train/The_thao.muc"]
	dev_files = ["../data_conll_topic/News_Dev/The_thao.muc"]
	test_files = ["../data_conll_topic/News_Test/The_thao.muc"]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 3
0
def xahoi_thethao_transfer(w):
	print("xahoi_thethao_transfer")
	train_files = [("../data_conll_topic/News_Train/Xa_hoi.muc", 0),
		("../data_conll_topic/News_Train/The_thao.muc", 1)]
	dev_files = [("../data_conll_topic/News_Dev/The_thao.muc", 1)]
	test_files = [("../data_conll_topic/News_Test/The_thao.muc", 1)]
	ner.run(train_files, dev_files, test_files, 500, weigth = w)
Ejemplo n.º 4
0
def kinhte_vanhoa_transfer_weigth(w):
	print("kinhte_vanhoa_transfer_no_shuffle")
	train_files = [("../data_conll_topic/News_Train/Kinh_te.muc", 0),
		("../data_conll_topic/News_Train/Van_hoa.muc", 1)]
	dev_files = [("../data_conll_topic/News_Dev/Van_hoa.muc", 1)]
	test_files = [("../data_conll_topic/News_Test/Van_hoa.muc", 1)]
	ner.run(train_files, dev_files, test_files, 500, weigth = w)
Ejemplo n.º 5
0
def kinhte_thethao_transfer_no_shuffle2():
	print("kinhte_thethao_transfer_no_shuffle2")
	train_files = [("../data_conll_topic/News_Train/The_thao.muc", 0),
		("../data_conll_topic/News_Train/Kinh_te.muc", 1)]
	dev_files = [("../data_conll_topic/News_Dev/The_thao.muc", 1)]
	test_files = [("../data_conll_topic/News_Test/The_thao.muc", 1)]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 6
0
def kinhte_vanhoa_mix_no_shuffle():
	print("kinhte_vanhoa_mix_no_shuffle")
	train_files = ["../data_conll_topic/News_Train/Kinh_te.muc",
		"../data_conll_topic/News_Train/Van_hoa.muc"]
	dev_files = ["../data_conll_topic/News_Dev/Van_hoa.muc"]
	test_files = ["../data_conll_topic/News_Test/Van_hoa.muc"]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 7
0
def thethao_kinhte_mix():
	train_files = ["../folds/fold_1/Kinh_te.train",
		"../folds/fold_1/The_thao.train",
		"../folds/fold_1/The_thao.dev",
		"../folds/fold_1/The_thao.test"]
	dev_files = ["../folds/fold_1/Kinh_te.dev"]
	test_files = ["../folds/fold_1/Kinh_te.test"]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 8
0
def kinhte_thethao_transfer():
	train_files = [("../folds/fold_1/The_thao.train", 1),
		("../folds/fold_1/Kinh_te.train", 0),
		("../folds/fold_1/Kinh_te.dev", 0),
		("../folds/fold_1/Kinh_te.test", 0)]
	dev_files = [("../folds/fold_1/The_thao.dev", 1)]
	test_files = [("../folds/fold_1/The_thao.test", 1)]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 9
0
def t1(): # old
	topics = ["Doi_song","Giao_duc","Kinh_te","The_gioi",\
		"Van_hoa","Giai_tri","KH-CN","Phap_luat","The_thao","Xa_hoi"]
	train_files = []
	test_files = []
	for topic in topics:
		train_files.append("../data_conll_topic/Train/%s.muc" % topic)
		test_files.append("../data_conll_topic/Dev/%s.muc" % topic)
	ner.run(train_files, test_files, 50)
Ejemplo n.º 10
0
def all_no_shuffle():
	print("all_no_shuffle")
	train_files = []
	dev_files = []
	test_files = []
	topics = ["Doi_song","Giao_duc","Kinh_te","The_gioi",\
		"Van_hoa","Giai_tri","KH-CN","Phap_luat","The_thao","Xa_hoi"]
	for topic in topics:
		train_files.append("../data_conll_topic/News_Train/%s.muc" % topic)
		dev_files.append("../data_conll_topic/News_Dev/%s.muc" % topic)
		test_files.append("../data_conll_topic/News_Test/%s.muc" % topic)
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 11
0
def social_1():
	# topics = ["Cong_nghe","Giao_duc","Nha_dat","The_thao","Doi_song","Khoa_hoc","Phap_luat","Van_hoa","Giai_tri","Kinh_te","The_gioi","Xa_hoi"]
	topics = ["Cong_nghe","Giao_duc","Nha_dat","The_thao","Doi_song","Khoa_hoc","Phap_luat"]
	# topics = ["Cong_nghe", "Kinh_te"]
	train_files = []
	dev_files = []
	test_files = []
	for topic in topics:
		train_files.append("../data_conll_topic/Social_Train/%s.muc" % topic)
		dev_files.append("../data_conll_topic/Social_Dev/%s.muc" % topic)
		test_files.append("../data_conll_topic/Social_Test/%s.muc" % topic)
	ner.run(train_files, dev_files, test_files, 100)
Ejemplo n.º 12
0
def news_1_no_val():
	# topics = ["Doi_song","Giao_duc","Kinh_te","The_gioi",\
	# 	"Van_hoa","Giai_tri","KH-CN","Phap_luat","The_thao","Xa_hoi"]
	topics = ["KH-CN", "Kinh_te"]
	train_files = []
	dev_files = []
	test_files = []
	for topic in topics:
		train_files.append("../data_conll_topic/News_Train/%s.muc" % topic)
		dev_files.append("../data_conll_topic/News_Dev/%s.muc" % topic)
		test_files.append("../data_conll_topic/News_Test/%s.muc" % topic)
	ner.run(train_files, dev_files, test_files, 50, no_val=True)
Ejemplo n.º 13
0
def kinh_te_2_lstm():
	train_files = ["../folds/fold_1/Kinh_te.train"]
	dev_files = ["../folds/fold_1/Kinh_te.dev"]
	test_files = ["../folds/fold_1/Kinh_te.test"]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 14
0
def kinhte_transfer(src):
	train_files = [("../data_conll_topic/News_Train/%s.muc" % src, 0),
		("../data_conll_topic/News_Train/Kinh_te.muc", 1)]
	dev_files = [("../data_conll_topic/News_Dev/Kinh_te.muc", 1)]
	test_files = [("../data_conll_topic/News_Test/Kinh_te.muc", 1)]
	ner.run(train_files, dev_files, test_files, 500, name="%s_kinhte_w_0.3", weigth = 0.3)
Ejemplo n.º 15
0
def kinhte_mix(src):
	train_files = ["../data_conll_topic/News_Train/%s.muc" % src,
		"../data_conll_topic/News_Train/Kinh_te.muc"]
	dev_files = ["../data_conll_topic/News_Dev/Kinh_te.muc"]
	test_files = ["../data_conll_topic/News_Test/Kinh_te.muc"]
	ner.run(train_files, dev_files, test_files, 500, name="%s_kinhte_mix")
Ejemplo n.º 16
0
def vanhoa_kinhte_indomain(in_domain):
	train_files = [	"../data_conll_topic/News_Train/Kinh_te.muc"]
	dev_files = ["../data_conll_topic/News_Dev/Kinh_te.muc"]
	test_files = ["../data_conll_topic/News_Test/Kinh_te.muc"]
	ner.run(train_files, dev_files, test_files, 500, in_domain=in_domain)
Ejemplo n.º 17
0
def vanhoa_kinhte_transfer():
	train_files = [("../data_conll_topic/News_Train/Van_hoa.muc", 0),
		("../data_conll_topic/News_Train/Kinh_te.muc", 1)]
	dev_files = [("../data_conll_topic/News_Dev/Kinh_te.muc", 1)]
	test_files = [("../data_conll_topic/News_Test/Kinh_te.muc", 1)]
	ner.run(train_files, dev_files, test_files, 500, name="vanhoa_kinhte_w_0", weigth = 0.0)
Ejemplo n.º 18
0
def kinh_te_transfer_model():
	train_files = ["../folds/fold_1/Kinh_te.train"]
	dev_files = ["../folds/fold_1/Kinh_te.dev"]
	test_files = ["../folds/fold_1/Kinh_te.test"]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 19
0
def topic_2_lstm(topic):
	train_files = ["../folds/fold_1/%s.train" % topic]
	dev_files = ["../folds/fold_1/%s.dev" % topic]
	test_files = ["../folds/fold_1/%s.test" % topic]
	ner.run(train_files, dev_files, test_files, 500, name=topic)
Ejemplo n.º 20
0
def kinhte_thethao_outdomain():
	train_files = ["../folds/fold_1/Kinh_te.train"]
	dev_files = ["../folds/fold_1/The_thao.dev"]
	test_files = ["../folds/fold_1/The_thao.test"]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 21
0
def thethao_xahoi_out():
	print("thethao_xahoi_out")
	train_files = ["../data_conll_topic/News_Train/The_thao.muc"]
	dev_files = ["../data_conll_topic/News_Dev/Xa_hoi.muc"]
	test_files = ["../data_conll_topic/News_Test/Xa_hoi.muc"]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 22
0
def vanhoa_kinhte_outdomain():
	print("vanhoa_kinhte_outdomain")
	train_files = ["../data_conll_topic/News_Train/Van_hoa.muc"]
	dev_files = ["../data_conll_topic/News_Dev/Kinh_te.muc"]
	test_files = ["../data_conll_topic/News_Test/Kinh_te.muc"]
	ner.run(train_files, dev_files, test_files, 500)	
Ejemplo n.º 23
0
def vanhoa_kinhte_transfer_no_shuffle():
	train_files = [("../data_conll_topic/News_Train/Van_hoa.muc", 0),
		("../data_conll_topic/News_Train/Kinh_te.muc", 1)]
	dev_files = [("../data_conll_topic/News_Dev/Kinh_te.muc", 1)]
	test_files = [("../data_conll_topic/News_Test/Kinh_te.muc", 2)]
	ner.run(train_files, dev_files, test_files, 500)
Ejemplo n.º 24
0
def thethao_2_lstm():
	print("thethao_2_lstm")
	train_files = ["../folds/fold_1/Kinh_te.train"]
	dev_files = ["../folds/fold_1/Kinh_te.dev"]
	test_files = ["../folds/fold_1/Kinh_te.test"]
	ner.run(train_files, dev_files, test_files, 500)