In [8]:
import os
os.chdir(r"C:\Users\Xiaozhou\Documents\Address")
import febrl as fe
import pandas as pd
import json
import random as rn

In [2]:
tags_dict = json.load(open("Tags.json"))
reg_subs=((";"," , "),(","," , "),("  "," "))
reg_tags=(("^,$","CM"),
("^no.$","NO"),
('^[0-9]+(rd|st|th)$',"TH"),
('^[0-9]+-[0-9]+[A-Za-z]*$',"RN"),
('^[0-9]$',"II"),
('^[0-9]+$',"NU"),
("UN"))

In [3]:
a=fe.standardiser()
a.load_tags(tags_dict)
a.load_reg_subs(reg_subs)
a.load_reg_tags(reg_tags)

In [4]:
df=pd.read_csv('Address.csv')

In [19]:
df.head(10)


Out[19]:
BuildingID AddressEng AddressTrad AddressSimp
0 1.0 HING SING BUILDING,NO. 8-8A,UN CHAU STREET,SHA... 九龍深水埗區元州街8-8A號慶成樓 九龙深水埗区元州街8-8A号庆成楼
1 2.0 HOUSE 36,FAIRVIEW PARK,9TH STREET,RIVER NORTH,... 新界元朗區河北第九街錦綉花園36號 新界元朗区河北第九街锦綉花园36号
2 2.0 FAIRVIEW PARK,NO. 36,FAIRVIEW PARK RIVER NORTH... 新界元朗區錦綉花園河北第九街36號錦綉花園 新界元朗区锦綉花园河北第九街36号锦綉花园
3 2.0 FAIRVIEW PARK,NO. 36,FAIRVIEW PARK 9TH STREET ... 新界元朗區錦綉花園河北第九街36號錦綉花園 新界元朗区锦綉花园河北第九街36号锦綉花园
4 3.0 NO. 51,SOUTH WALL ROAD,KOWLOON CITY,KOWLOON 九龍九龍城區城南道51號 九龙九龙城区城南道51号
5 4.0 HOUSE 20,FAIRVIEW PARK SECTION C, 1ST STREET,Y... 新界元朗區C段第一街錦綉花園20號 新界元朗区C段第一街锦綉花园20号
6 4.0 FAIRVIEW PARK,NO. 20,FAIRVIEW PARK SECTION C; ... 新界元朗區錦綉花園C段第一街20號錦綉花園 新界元朗区锦綉花园C段第一街20号锦綉花园
7 4.0 FAIRVIEW PARK,NO. 20,FAIRVIEW PARK SECT C 1ST ... 新界元朗區錦綉花園C段第一街20號錦綉花園 新界元朗区锦綉花园C段第一街20号锦綉花园
8 5.0 HARVEST BUILDING,NO. 29-35,WING KUT STREET,CEN... 香港島中西區永吉街29-35號恒豐大廈 香港岛中西区永吉街29-35号恒丰大厦
9 6.0 WAI TUNG HOUSE,TUNG TAU (II) ESTATE,NO. 183,TU... 九龍黃大仙區東頭村道183號東頭(二)邨偉東樓 九龙黄大仙区东头村道183号东头(二)邨伟东楼

In [6]:
data=fe.read_train_data('hk_add_train.csv')
a.train('asdf', data, 'laplace')


WARNING:root:HMM state "area" has transition probabilities sum not 1.0: 0.000000

In [18]:
x=rn.randrange(0,65000)
text=df['AddressEng'][x]
print(x)
print(text)
print(a.tag_str(text))
a.std_str(text)


2191
NO. 278-280,KI LUNG STREET,SHAM SHUI PO,KOWLOON
('no. 278-280 , ki lung street , sham_shui_po , kowloon', 'NO RN CM UN UN RD CM A2/DT CM A1')
Out[18]:
{'area': 'kowloon',
 'com': ', , ,',
 'street': 'ki lung street',
 'strnum': 'no. 278-280',
 'subarea': 'sham shui po'}