-
Notifications
You must be signed in to change notification settings - Fork 0
/
clustering.py
171 lines (139 loc) · 6.05 KB
/
clustering.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
import gensim
import re
import numpy as np
from sklearn.cluster import AffinityPropagation
from sklearn import metrics
from collections import defaultdict
from scipy import spatial
import sys
import json
def build_matrix(model, word_list, counts):
matrix = np.zeros((len(word_list), len(word_list)))
for i in range(len(word_list)):
for j in range(i, len(word_list)):
if i == j:
matrix[i,j] = 1.0
elif word_list[i].startswith(word_list[j]) or word_list[j].startswith(word_list[i]):
matrix[i,j] = matrix[j,i] = 0.99
elif (lang in ['sv', 'svl', 'svk'] and
(word_list[i][0:10] == 'rheumatism' and word_list[j][0:9] == 'reumatism' or
word_list[j][0:10] == 'rheumatism' and word_list[i][0:9] == 'reumatism')):
matrix[i,j] = matrix[j,i] = 0.99
elif (lang in ['sv', 'svl', 'svk'] and
(word_list[i][0:11] == 'katholicism' and word_list[j][0:10] == 'katolicism' or
word_list[j][0:11] == 'katholicism' and word_list[i][0:10] == 'katolicism')):
matrix[i,j] = matrix[j,i] = 0.99
else:
matrix[i,j] = matrix[j,i] = (1-spatial.distance.cosine(model[word_list[i]], model[word_list[j]]))
return matrix
def cluster(model, prev_model):
model.init_sims(replace=True)
ism = {}
for k,v in model.wv.vocab.items():
if lang in ['sv', 'svl', 'svk']:
regexp = "ism$|ismen$|ismens$"
else:
regexp = "ismi$|ismin$|ismia$|ismissa$|ismista$|ismilla$|ismille$|ismilta$|ismina$|ismiksi$|ismitta$|ismit$|ismien$|ismeja$|ismeissa$|ismeista$|ismeihin$|ismeilla$|ismeilta$|ismille$|ismeina$|ismeiksi$|ismein$|ismeitta$|ismeineen$|ismiä$|ismissä$|ismistä$|ismillä$|ismiltä$|isminä$|ismittä$|ismejä$|ismeissä$|ismeistä$|ismeillä$|ismeiltä$|ismeinä$|ismeittä$"
min_len = 6 if lang in ["fi", "fil"] else 5
if len(k) >= min_len and re.search(regexp, k):
count = v.count
if k in prev_model.wv.vocab:
count = count - prev_model.wv.vocab[k].count
if count > 0:
ism[k] = count
word_list = list(ism)
if len(word_list) == 0:
return None
matrix = build_matrix(model, word_list, ism)
af = AffinityPropagation(affinity="precomputed", verbose=False).fit(matrix)
return (word_list, ism, af.labels_, af.cluster_centers_indices_)
def print_cluster(word_list, counts, labels, centers):
cluster = defaultdict(list)
center = {}
for i in range(len(labels)):
if i in centers:
center[labels[i]] = word_list[i]
cluster[labels[i]].append(word_list[i])
for c in sorted(cluster, key = lambda x: len(cluster[x]), reverse=True):
cl = cluster[c]
print (center[c], counts[center[c]])
words = {w:count for w,count in counts.items() if w in cl}
for w in sorted(words, key=words.get, reverse=True):
if not w==center[c]:
print (w, words[w])
print ("\n")
def make_json(word_list, counts, labels, centers):
# 1 most frequent
outlist = []
# 2 most frequent
outlist2 = []
# most freq + centroid
outlist_cf = []
cluster = defaultdict(list)
center = {}
for i in range(len(labels)):
if i in centers:
center[labels[i]] = word_list[i]
cluster[labels[i]].append(word_list[i])
for c in sorted(cluster, key = lambda x: len(cluster[x]), reverse=True):
cl = cluster[c]
words = {w:count for w,count in counts.items() if w in cl}
res = [w for w in sorted(words, key=words.get, reverse=True)]
if len(res) > 2:
res2 = [res[0]+"_"+res[1]] + res[2:]
res_noc = [r for r in res if r != center[c]]
res_cf = [center[c] + "_" + res_noc[0]] + res_noc[2:]
elif len(res) == 2:
res2 = res_cf = [res[0] + "_" + res[1]]
else:
res2 = res_cf = res
outlist.append(res)
outlist2.append(res2)
outlist_cf.append(res_cf)
return outlist, outlist2, outlist_cf
if __name__ == "__main__":
try:
lang = sys.argv[1]
assert(lang in ['fi', 'fil', 'sv', 'svl', 'svk'])
except:
print("usage: clustering.py [fi|fil|sv|svl|svk] <output_json_path>")
exit(1)
if lang == 'fi':
ys = ['1760', '1820', '1840', '1860', '1880', '1900']
base_path = "../models/FI_models/model_fi_"
elif lang == 'fil':
ys = ['1760', '1820', '1840', '1860', '1880', '1900']
base_path = "../models/FI_lemma/model_fi_"
elif lang == 'sv':
ys = ['1740', '1760', '1780', '1800', '1820', '1840', '1860', '1880', '1900']
base_path = "../models/SV_out_new/model_sv_"
elif lang == 'svl':
# ys = ['1740', '1760', '1780', '1800', '1820', '1840', '1860', '1880', '1900']
ys = ['1760', '1780', '1800', '1820', '1840', '1860', '1880']
base_path = "../models/SV_lowercase/model_sv_"
elif lang == 'svk':
ys = ['1740', '1760', '1780', '1800', '1820', '1840', '1860', '1880', '1900']
base_path = "../models/SV_diachronic/models/model_sv_"
prev_model = None
res_dict = {}
res_dict2 = {}
res_dict_cf = {}
for y in ys:
model_path = base_path +y+".w2v"
model = gensim.models.Word2Vec.load(model_path)
clustering = cluster(model, prev_model)
print("\n*********************\n")
print(y)
if clustering is None:
print("nothing")
else:
print_cluster(*clustering)
res_dict[y], res_dict2[y], res_dict_cf[y] = make_json(*clustering)
prev_model = model
jp, jp2, jp_cf = 'clustering.json', 'clustering2.json', 'clustering_cf.json'
with open(jp, 'w') as jout:
json.dump(res_dict, jout)
with open(jp2, 'w') as jout:
json.dump(res_dict2, jout)
with open(jp_cf, 'w') as jout:
json.dump(res_dict_cf, jout)