Created
May 23, 2016 11:35
-
-
Save farizrahman4u/c973821e70a850d5dd5aaa6e522e3889 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import numpy as np | |
__author__ = 'Fariz Rahman' | |
def eq(x, y): | |
return x.lower().replace(" ", "") == y.lower().replace(" ", "") | |
def get_words(x): | |
x = x.replace(" ", " ") | |
words = x.split(" ") | |
if "" in words: | |
words.remove("") | |
return words | |
def match(nb1, nb2): | |
l_1 = set(nb1[0]) | |
l_2 = set(nb2[0]) | |
n_l1 = len(l_1) | |
n_l2 = len(l_2) | |
if n_l1 + n_l2 == 0: | |
l_score = .5 | |
else: | |
n_l_intersect = n_l1 - len(l_1 - l_2) | |
l_score = float(n_l_intersect) / (n_l1 + n_l2) | |
r_1 = set(nb1[0]) | |
r_2 = set(nb2[0]) | |
n_r1 = len(r_1) | |
n_r2 = len(r_2) | |
if n_r1 + n_r2 == 0: | |
r_score = .5 | |
else: | |
n_r_intersect = n_r1 - len(r_1 - r_2) | |
r_score = float(n_r_intersect) / (n_r1 + n_r2) | |
return l_score + r_score | |
def get_neighbourhood(s, x): | |
left = [] | |
right = [] | |
n = len(s) | |
for i in range(n): | |
w = s[i] | |
if eq(w, x): | |
left += s[0 : i] | |
right += s[i + 1 : n] | |
return [left, right] | |
def get_transformation(q, a): | |
qw = get_words(q) | |
aw = get_words(a) | |
transform = [] | |
for w1 in aw: | |
flag = False | |
for w2 in qw: | |
if eq(w1, w2): | |
flag = w2 | |
break | |
if flag: | |
elem = ['neighbourhood', get_neighbourhood(qw, flag)] | |
else: | |
elem = ['word', w1] | |
transform += [elem] | |
return transform | |
def get_all_neighbourhoods(s): | |
nbs = [] | |
for w in s: | |
nbs += [get_neighbourhood(s, w)] | |
return nbs | |
def get_max_match(nb, nbs): | |
scores = [] | |
for nbx in nbs: | |
scores += [match(nbx, nb)] | |
return np.argmax(scores) | |
def apply_transformation(q, transformation): | |
a = "" | |
q = get_words(q) | |
nbs = get_all_neighbourhoods(q) | |
for elem in transformation: | |
if elem[0] == 'word': | |
a += elem[1] + ' ' | |
else: | |
a += q[get_max_match(elem[1], nbs)] + ' ' | |
return a[:-1] | |
Q = "please open the door" | |
A = "door : open" | |
t = get_transformation(Q, A) | |
Q2 = "close the window" | |
A2 = apply_transformation(Q2, t) | |
print A2 | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
thanks! would you talk about the general concept of how you're trying to do this?