raske-nlp 0.1.5
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +23 -0
- data/.travis.yml +5 -0
- data/Gemfile +6 -0
- data/Gemfile.lock +20 -0
- data/MIT-LICENSE +21 -0
- data/README.md +31 -0
- data/Rakefile +1 -0
- data/bin/console +14 -0
- data/data/AliStopList.txt +309 -0
- data/data/StopWordsIso.txt +733 -0
- data/lib/raske-nlp.rb +17 -0
- data/lib/raske-nlp/result.rb +79 -0
- data/lib/raske-nlp/stop_list.rb +22 -0
- data/lib/raske-nlp/version.rb +3 -0
- data/raske-nlp.gemspec +23 -0
- metadata +86 -0
checksums.yaml
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
1
|
+
---
|
|
2
|
+
SHA256:
|
|
3
|
+
metadata.gz: ecbcbf018652bd37d47dc9fa14976542a922a28a9e7a0fa6b5fd5433320591cf
|
|
4
|
+
data.tar.gz: cde1f719a9f497ab2973a0c6e66a23dc06142fe24c61d5c4bf4277b0ed918fba
|
|
5
|
+
SHA512:
|
|
6
|
+
metadata.gz: b633edb8e0bf49796a073f888526da8b3608a702180275858db399ac917e40b0e2fb96ed4d682814cf6e2a871121ea2371b1c94f5d4332b24a95659b69ec874b
|
|
7
|
+
data.tar.gz: 499dbc7271456165a6a272382adaf96da404d6226536817464d0848c8deaaea27196ce6e9432f9087e20bfc58571489812b1fb63cf4e0f495cd56c16be86cdcf
|
data/.gitignore
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
1
|
+
/.bundle/
|
|
2
|
+
/.yardoc
|
|
3
|
+
/_yardoc/
|
|
4
|
+
/coverage/
|
|
5
|
+
/doc/
|
|
6
|
+
/pkg/
|
|
7
|
+
/spec/reports/
|
|
8
|
+
/tmp/
|
|
9
|
+
|
|
10
|
+
# rspec failure tracking
|
|
11
|
+
.rspec_status
|
|
12
|
+
.ruby-version
|
|
13
|
+
.ruby-gemset
|
|
14
|
+
*.gem
|
|
15
|
+
|
|
16
|
+
|
|
17
|
+
result = RaskeNLP.run(text, {
|
|
18
|
+
min_phrase_length: 1,
|
|
19
|
+
max_phrase_length: 3,
|
|
20
|
+
min_frequency: 1,
|
|
21
|
+
min_score: 1,
|
|
22
|
+
stop_list: RaskeNLP::StopList::ALI
|
|
23
|
+
})
|
data/.travis.yml
ADDED
data/Gemfile
ADDED
data/Gemfile.lock
ADDED
data/MIT-LICENSE
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
1
|
+
The MIT License (MIT)
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2017 Sam Pohlenz
|
|
4
|
+
|
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
|
7
|
+
in the Software without restriction, including without limitation the rights
|
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
|
10
|
+
furnished to do so, subject to the following conditions:
|
|
11
|
+
|
|
12
|
+
The above copyright notice and this permission notice shall be included in
|
|
13
|
+
all copies or substantial portions of the Software.
|
|
14
|
+
|
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
|
21
|
+
THE SOFTWARE.
|
data/README.md
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
1
|
+
# raske-nlp
|
|
2
|
+
|
|
3
|
+
> Ruby implementation of the Rapid Automatic Spanish Keyword Extraction (RAKE) algorithm
|
|
4
|
+
|
|
5
|
+
## Usage
|
|
6
|
+
|
|
7
|
+
```ruby
|
|
8
|
+
text = "Uno de enero, dos de febrero, tres de marzo, cuatro de abril, cinco de mayo,
|
|
9
|
+
seis de junio siete de julio, ¡SAN FERMÍN! Uno de enero, dos de febrero,
|
|
10
|
+
tres de marzo, cuatro de abril, cinco de mayo, seis de junio siete de julio, ¡SAN FERMÍN!
|
|
11
|
+
A Pamplona hemos de ir, con una media, con una media, a Pamplona hemos de ir
|
|
12
|
+
con una media y un calcetín."
|
|
13
|
+
|
|
14
|
+
result = RaskeNLP.run(text, {
|
|
15
|
+
min_phrase_length: 1,
|
|
16
|
+
max_phrase_length: 3,
|
|
17
|
+
min_frequency: 1,
|
|
18
|
+
min_score: 1,
|
|
19
|
+
stop_list: RaskeNLP::StopList::ISO
|
|
20
|
+
})
|
|
21
|
+
|
|
22
|
+
result.keywords
|
|
23
|
+
```
|
|
24
|
+
|
|
25
|
+
```ruby
|
|
26
|
+
=> {"san fermín"=>15.333333333333334, "calcetín"=>5.333333333333334, "media"=>1.0, "pamplona"=>1.0, "julio"=>1.0, "junio"=>1.0, "mayo"=>1.0, "abril"=>1.0, "marzo"=>1.0, "febrero"=>1.0, "enero"=>1.0}
|
|
27
|
+
```
|
|
28
|
+
|
|
29
|
+
## License
|
|
30
|
+
|
|
31
|
+
The gem is available as open source under the terms of the [MIT License](https://opensource.org/licenses/MIT).
|
data/Rakefile
ADDED
|
@@ -0,0 +1 @@
|
|
|
1
|
+
require "bundler/gem_tasks"
|
data/bin/console
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
#!/usr/bin/env ruby
|
|
2
|
+
|
|
3
|
+
require "bundler/setup"
|
|
4
|
+
require "raske-nlp"
|
|
5
|
+
|
|
6
|
+
TEXT = <<-EOF.strip
|
|
7
|
+
Compatibility of systems of linear constraints over the set of natural numbers
|
|
8
|
+
Criteria of compatibility of a system of linear Diophantine equations, strict inequations, and nonstrict inequations are considered. Upper bounds for components of a minimal set of solutions and algorithms of construction of minimal generating sets of solutions for all types of systems are given. These criteria and the corresponding algorithms for constructing a minimal supporting set of solutions can be used in solving all the considered types of systems and systems of mixed types.
|
|
9
|
+
EOF
|
|
10
|
+
|
|
11
|
+
puts "TEXT: #{TEXT.inspect}"
|
|
12
|
+
|
|
13
|
+
require "irb"
|
|
14
|
+
IRB.start(__FILE__)
|
|
@@ -0,0 +1,309 @@
|
|
|
1
|
+
# From https://raw.githubusercontent.com/Alir3z4/stop-words/master/spanish.txt
|
|
2
|
+
a
|
|
3
|
+
al
|
|
4
|
+
algo
|
|
5
|
+
algunas
|
|
6
|
+
algunos
|
|
7
|
+
ante
|
|
8
|
+
antes
|
|
9
|
+
como
|
|
10
|
+
con
|
|
11
|
+
contra
|
|
12
|
+
cual
|
|
13
|
+
cuando
|
|
14
|
+
de
|
|
15
|
+
del
|
|
16
|
+
desde
|
|
17
|
+
donde
|
|
18
|
+
durante
|
|
19
|
+
e
|
|
20
|
+
el
|
|
21
|
+
ella
|
|
22
|
+
ellas
|
|
23
|
+
ellos
|
|
24
|
+
en
|
|
25
|
+
entre
|
|
26
|
+
era
|
|
27
|
+
erais
|
|
28
|
+
eran
|
|
29
|
+
eras
|
|
30
|
+
eres
|
|
31
|
+
es
|
|
32
|
+
esa
|
|
33
|
+
esas
|
|
34
|
+
ese
|
|
35
|
+
eso
|
|
36
|
+
esos
|
|
37
|
+
esta
|
|
38
|
+
estaba
|
|
39
|
+
estabais
|
|
40
|
+
estaban
|
|
41
|
+
estabas
|
|
42
|
+
estad
|
|
43
|
+
estada
|
|
44
|
+
estadas
|
|
45
|
+
estado
|
|
46
|
+
estados
|
|
47
|
+
estamos
|
|
48
|
+
estando
|
|
49
|
+
estar
|
|
50
|
+
estaremos
|
|
51
|
+
estará
|
|
52
|
+
estarán
|
|
53
|
+
estarás
|
|
54
|
+
estaré
|
|
55
|
+
estaréis
|
|
56
|
+
estaría
|
|
57
|
+
estaríais
|
|
58
|
+
estaríamos
|
|
59
|
+
estarían
|
|
60
|
+
estarías
|
|
61
|
+
estas
|
|
62
|
+
este
|
|
63
|
+
estemos
|
|
64
|
+
esto
|
|
65
|
+
estos
|
|
66
|
+
estoy
|
|
67
|
+
estuve
|
|
68
|
+
estuviera
|
|
69
|
+
estuvierais
|
|
70
|
+
estuvieran
|
|
71
|
+
estuvieras
|
|
72
|
+
estuvieron
|
|
73
|
+
estuviese
|
|
74
|
+
estuvieseis
|
|
75
|
+
estuviesen
|
|
76
|
+
estuvieses
|
|
77
|
+
estuvimos
|
|
78
|
+
estuviste
|
|
79
|
+
estuvisteis
|
|
80
|
+
estuviéramos
|
|
81
|
+
estuviésemos
|
|
82
|
+
estuvo
|
|
83
|
+
está
|
|
84
|
+
estábamos
|
|
85
|
+
estáis
|
|
86
|
+
están
|
|
87
|
+
estás
|
|
88
|
+
esté
|
|
89
|
+
estéis
|
|
90
|
+
estén
|
|
91
|
+
estés
|
|
92
|
+
fue
|
|
93
|
+
fuera
|
|
94
|
+
fuerais
|
|
95
|
+
fueran
|
|
96
|
+
fueras
|
|
97
|
+
fueron
|
|
98
|
+
fuese
|
|
99
|
+
fueseis
|
|
100
|
+
fuesen
|
|
101
|
+
fueses
|
|
102
|
+
fui
|
|
103
|
+
fuimos
|
|
104
|
+
fuiste
|
|
105
|
+
fuisteis
|
|
106
|
+
fuéramos
|
|
107
|
+
fuésemos
|
|
108
|
+
ha
|
|
109
|
+
habida
|
|
110
|
+
habidas
|
|
111
|
+
habido
|
|
112
|
+
habidos
|
|
113
|
+
habiendo
|
|
114
|
+
habremos
|
|
115
|
+
habrá
|
|
116
|
+
habrán
|
|
117
|
+
habrás
|
|
118
|
+
habré
|
|
119
|
+
habréis
|
|
120
|
+
habría
|
|
121
|
+
habríais
|
|
122
|
+
habríamos
|
|
123
|
+
habrían
|
|
124
|
+
habrías
|
|
125
|
+
habéis
|
|
126
|
+
había
|
|
127
|
+
habíais
|
|
128
|
+
habíamos
|
|
129
|
+
habían
|
|
130
|
+
habías
|
|
131
|
+
han
|
|
132
|
+
has
|
|
133
|
+
hasta
|
|
134
|
+
hay
|
|
135
|
+
haya
|
|
136
|
+
hayamos
|
|
137
|
+
hayan
|
|
138
|
+
hayas
|
|
139
|
+
hayáis
|
|
140
|
+
he
|
|
141
|
+
hemos
|
|
142
|
+
hube
|
|
143
|
+
hubiera
|
|
144
|
+
hubierais
|
|
145
|
+
hubieran
|
|
146
|
+
hubieras
|
|
147
|
+
hubieron
|
|
148
|
+
hubiese
|
|
149
|
+
hubieseis
|
|
150
|
+
hubiesen
|
|
151
|
+
hubieses
|
|
152
|
+
hubimos
|
|
153
|
+
hubiste
|
|
154
|
+
hubisteis
|
|
155
|
+
hubiéramos
|
|
156
|
+
hubiésemos
|
|
157
|
+
hubo
|
|
158
|
+
la
|
|
159
|
+
las
|
|
160
|
+
le
|
|
161
|
+
les
|
|
162
|
+
lo
|
|
163
|
+
los
|
|
164
|
+
me
|
|
165
|
+
mi
|
|
166
|
+
mis
|
|
167
|
+
mucho
|
|
168
|
+
muchos
|
|
169
|
+
muy
|
|
170
|
+
más
|
|
171
|
+
mí
|
|
172
|
+
mía
|
|
173
|
+
mías
|
|
174
|
+
mío
|
|
175
|
+
míos
|
|
176
|
+
nada
|
|
177
|
+
ni
|
|
178
|
+
no
|
|
179
|
+
nos
|
|
180
|
+
nosotras
|
|
181
|
+
nosotros
|
|
182
|
+
nuestra
|
|
183
|
+
nuestras
|
|
184
|
+
nuestro
|
|
185
|
+
nuestros
|
|
186
|
+
o
|
|
187
|
+
os
|
|
188
|
+
otra
|
|
189
|
+
otras
|
|
190
|
+
otro
|
|
191
|
+
otros
|
|
192
|
+
para
|
|
193
|
+
pero
|
|
194
|
+
poco
|
|
195
|
+
por
|
|
196
|
+
porque
|
|
197
|
+
que
|
|
198
|
+
quien
|
|
199
|
+
quienes
|
|
200
|
+
qué
|
|
201
|
+
se
|
|
202
|
+
sea
|
|
203
|
+
seamos
|
|
204
|
+
sean
|
|
205
|
+
seas
|
|
206
|
+
seremos
|
|
207
|
+
será
|
|
208
|
+
serán
|
|
209
|
+
serás
|
|
210
|
+
seré
|
|
211
|
+
seréis
|
|
212
|
+
sería
|
|
213
|
+
seríais
|
|
214
|
+
seríamos
|
|
215
|
+
serían
|
|
216
|
+
serías
|
|
217
|
+
seáis
|
|
218
|
+
sido
|
|
219
|
+
siendo
|
|
220
|
+
sin
|
|
221
|
+
sobre
|
|
222
|
+
sois
|
|
223
|
+
somos
|
|
224
|
+
son
|
|
225
|
+
soy
|
|
226
|
+
su
|
|
227
|
+
sus
|
|
228
|
+
suya
|
|
229
|
+
suyas
|
|
230
|
+
suyo
|
|
231
|
+
suyos
|
|
232
|
+
sí
|
|
233
|
+
también
|
|
234
|
+
tanto
|
|
235
|
+
te
|
|
236
|
+
tendremos
|
|
237
|
+
tendrá
|
|
238
|
+
tendrán
|
|
239
|
+
tendrás
|
|
240
|
+
tendré
|
|
241
|
+
tendréis
|
|
242
|
+
tendría
|
|
243
|
+
tendríais
|
|
244
|
+
tendríamos
|
|
245
|
+
tendrían
|
|
246
|
+
tendrías
|
|
247
|
+
tened
|
|
248
|
+
tenemos
|
|
249
|
+
tenga
|
|
250
|
+
tengamos
|
|
251
|
+
tengan
|
|
252
|
+
tengas
|
|
253
|
+
tengo
|
|
254
|
+
tengáis
|
|
255
|
+
tenida
|
|
256
|
+
tenidas
|
|
257
|
+
tenido
|
|
258
|
+
tenidos
|
|
259
|
+
teniendo
|
|
260
|
+
tenéis
|
|
261
|
+
tenía
|
|
262
|
+
teníais
|
|
263
|
+
teníamos
|
|
264
|
+
tenían
|
|
265
|
+
tenías
|
|
266
|
+
ti
|
|
267
|
+
tiene
|
|
268
|
+
tienen
|
|
269
|
+
tienes
|
|
270
|
+
todo
|
|
271
|
+
todos
|
|
272
|
+
tu
|
|
273
|
+
tus
|
|
274
|
+
tuve
|
|
275
|
+
tuviera
|
|
276
|
+
tuvierais
|
|
277
|
+
tuvieran
|
|
278
|
+
tuvieras
|
|
279
|
+
tuvieron
|
|
280
|
+
tuviese
|
|
281
|
+
tuvieseis
|
|
282
|
+
tuviesen
|
|
283
|
+
tuvieses
|
|
284
|
+
tuvimos
|
|
285
|
+
tuviste
|
|
286
|
+
tuvisteis
|
|
287
|
+
tuviéramos
|
|
288
|
+
tuviésemos
|
|
289
|
+
tuvo
|
|
290
|
+
tuya
|
|
291
|
+
tuyas
|
|
292
|
+
tuyo
|
|
293
|
+
tuyos
|
|
294
|
+
tú
|
|
295
|
+
un
|
|
296
|
+
una
|
|
297
|
+
uno
|
|
298
|
+
unos
|
|
299
|
+
vosotras
|
|
300
|
+
vosotros
|
|
301
|
+
vuestra
|
|
302
|
+
vuestras
|
|
303
|
+
vuestro
|
|
304
|
+
vuestros
|
|
305
|
+
y
|
|
306
|
+
ya
|
|
307
|
+
yo
|
|
308
|
+
él
|
|
309
|
+
éramos
|
|
@@ -0,0 +1,733 @@
|
|
|
1
|
+
#stop word list from ISO (Salton,1971). Available at https://github.com/stopwords-iso/stopwords-es/blob/master/stopwords-es.txt
|
|
2
|
+
0
|
|
3
|
+
1
|
|
4
|
+
2
|
|
5
|
+
3
|
|
6
|
+
4
|
|
7
|
+
5
|
|
8
|
+
6
|
|
9
|
+
7
|
|
10
|
+
8
|
|
11
|
+
9
|
|
12
|
+
_
|
|
13
|
+
a
|
|
14
|
+
actualmente
|
|
15
|
+
acuerdo
|
|
16
|
+
adelante
|
|
17
|
+
ademas
|
|
18
|
+
además
|
|
19
|
+
adrede
|
|
20
|
+
afirmó
|
|
21
|
+
agregó
|
|
22
|
+
ahi
|
|
23
|
+
ahora
|
|
24
|
+
ahí
|
|
25
|
+
al
|
|
26
|
+
algo
|
|
27
|
+
alguna
|
|
28
|
+
algunas
|
|
29
|
+
alguno
|
|
30
|
+
algunos
|
|
31
|
+
algún
|
|
32
|
+
alli
|
|
33
|
+
allí
|
|
34
|
+
alrededor
|
|
35
|
+
ambos
|
|
36
|
+
ampleamos
|
|
37
|
+
antano
|
|
38
|
+
antaño
|
|
39
|
+
ante
|
|
40
|
+
anterior
|
|
41
|
+
antes
|
|
42
|
+
apenas
|
|
43
|
+
aproximadamente
|
|
44
|
+
aquel
|
|
45
|
+
aquella
|
|
46
|
+
aquellas
|
|
47
|
+
aquello
|
|
48
|
+
aquellos
|
|
49
|
+
aqui
|
|
50
|
+
aquél
|
|
51
|
+
aquélla
|
|
52
|
+
aquéllas
|
|
53
|
+
aquéllos
|
|
54
|
+
aquí
|
|
55
|
+
arriba
|
|
56
|
+
arribaabajo
|
|
57
|
+
aseguró
|
|
58
|
+
asi
|
|
59
|
+
así
|
|
60
|
+
atras
|
|
61
|
+
aun
|
|
62
|
+
aunque
|
|
63
|
+
ayer
|
|
64
|
+
añadió
|
|
65
|
+
aún
|
|
66
|
+
b
|
|
67
|
+
bajo
|
|
68
|
+
bastante
|
|
69
|
+
bien
|
|
70
|
+
breve
|
|
71
|
+
buen
|
|
72
|
+
buena
|
|
73
|
+
buenas
|
|
74
|
+
bueno
|
|
75
|
+
buenos
|
|
76
|
+
c
|
|
77
|
+
cada
|
|
78
|
+
casi
|
|
79
|
+
cerca
|
|
80
|
+
cierta
|
|
81
|
+
ciertas
|
|
82
|
+
cierto
|
|
83
|
+
ciertos
|
|
84
|
+
cinco
|
|
85
|
+
claro
|
|
86
|
+
comentó
|
|
87
|
+
como
|
|
88
|
+
con
|
|
89
|
+
conmigo
|
|
90
|
+
conocer
|
|
91
|
+
conseguimos
|
|
92
|
+
conseguir
|
|
93
|
+
considera
|
|
94
|
+
consideró
|
|
95
|
+
consigo
|
|
96
|
+
consigue
|
|
97
|
+
consiguen
|
|
98
|
+
consigues
|
|
99
|
+
contigo
|
|
100
|
+
contra
|
|
101
|
+
cosas
|
|
102
|
+
creo
|
|
103
|
+
cual
|
|
104
|
+
cuales
|
|
105
|
+
cualquier
|
|
106
|
+
cuando
|
|
107
|
+
cuanta
|
|
108
|
+
cuantas
|
|
109
|
+
cuanto
|
|
110
|
+
cuantos
|
|
111
|
+
cuatro
|
|
112
|
+
cuenta
|
|
113
|
+
cuál
|
|
114
|
+
cuáles
|
|
115
|
+
cuándo
|
|
116
|
+
cuánta
|
|
117
|
+
cuántas
|
|
118
|
+
cuánto
|
|
119
|
+
cuántos
|
|
120
|
+
cómo
|
|
121
|
+
d
|
|
122
|
+
da
|
|
123
|
+
dado
|
|
124
|
+
dan
|
|
125
|
+
dar
|
|
126
|
+
de
|
|
127
|
+
debajo
|
|
128
|
+
debe
|
|
129
|
+
deben
|
|
130
|
+
debido
|
|
131
|
+
decir
|
|
132
|
+
dejó
|
|
133
|
+
del
|
|
134
|
+
delante
|
|
135
|
+
demasiado
|
|
136
|
+
demás
|
|
137
|
+
dentro
|
|
138
|
+
deprisa
|
|
139
|
+
desde
|
|
140
|
+
despacio
|
|
141
|
+
despues
|
|
142
|
+
después
|
|
143
|
+
detras
|
|
144
|
+
detrás
|
|
145
|
+
dia
|
|
146
|
+
dias
|
|
147
|
+
dice
|
|
148
|
+
dicen
|
|
149
|
+
dicho
|
|
150
|
+
dieron
|
|
151
|
+
diferente
|
|
152
|
+
diferentes
|
|
153
|
+
dijeron
|
|
154
|
+
dijo
|
|
155
|
+
dio
|
|
156
|
+
donde
|
|
157
|
+
dos
|
|
158
|
+
durante
|
|
159
|
+
día
|
|
160
|
+
días
|
|
161
|
+
dónde
|
|
162
|
+
e
|
|
163
|
+
ejemplo
|
|
164
|
+
el
|
|
165
|
+
ella
|
|
166
|
+
ellas
|
|
167
|
+
ello
|
|
168
|
+
ellos
|
|
169
|
+
embargo
|
|
170
|
+
empleais
|
|
171
|
+
emplean
|
|
172
|
+
emplear
|
|
173
|
+
empleas
|
|
174
|
+
empleo
|
|
175
|
+
en
|
|
176
|
+
encima
|
|
177
|
+
encuentra
|
|
178
|
+
enfrente
|
|
179
|
+
enseguida
|
|
180
|
+
entonces
|
|
181
|
+
entre
|
|
182
|
+
era
|
|
183
|
+
erais
|
|
184
|
+
eramos
|
|
185
|
+
eran
|
|
186
|
+
eras
|
|
187
|
+
eres
|
|
188
|
+
es
|
|
189
|
+
esa
|
|
190
|
+
esas
|
|
191
|
+
ese
|
|
192
|
+
eso
|
|
193
|
+
esos
|
|
194
|
+
esta
|
|
195
|
+
estaba
|
|
196
|
+
estabais
|
|
197
|
+
estaban
|
|
198
|
+
estabas
|
|
199
|
+
estad
|
|
200
|
+
estada
|
|
201
|
+
estadas
|
|
202
|
+
estado
|
|
203
|
+
estados
|
|
204
|
+
estais
|
|
205
|
+
estamos
|
|
206
|
+
estan
|
|
207
|
+
estando
|
|
208
|
+
estar
|
|
209
|
+
estaremos
|
|
210
|
+
estará
|
|
211
|
+
estarán
|
|
212
|
+
estarás
|
|
213
|
+
estaré
|
|
214
|
+
estaréis
|
|
215
|
+
estaría
|
|
216
|
+
estaríais
|
|
217
|
+
estaríamos
|
|
218
|
+
estarían
|
|
219
|
+
estarías
|
|
220
|
+
estas
|
|
221
|
+
este
|
|
222
|
+
estemos
|
|
223
|
+
esto
|
|
224
|
+
estos
|
|
225
|
+
estoy
|
|
226
|
+
estuve
|
|
227
|
+
estuviera
|
|
228
|
+
estuvierais
|
|
229
|
+
estuvieran
|
|
230
|
+
estuvieras
|
|
231
|
+
estuvieron
|
|
232
|
+
estuviese
|
|
233
|
+
estuvieseis
|
|
234
|
+
estuviesen
|
|
235
|
+
estuvieses
|
|
236
|
+
estuvimos
|
|
237
|
+
estuviste
|
|
238
|
+
estuvisteis
|
|
239
|
+
estuviéramos
|
|
240
|
+
estuviésemos
|
|
241
|
+
estuvo
|
|
242
|
+
está
|
|
243
|
+
estábamos
|
|
244
|
+
estáis
|
|
245
|
+
están
|
|
246
|
+
estás
|
|
247
|
+
esté
|
|
248
|
+
estéis
|
|
249
|
+
estén
|
|
250
|
+
estés
|
|
251
|
+
ex
|
|
252
|
+
excepto
|
|
253
|
+
existe
|
|
254
|
+
existen
|
|
255
|
+
explicó
|
|
256
|
+
expresó
|
|
257
|
+
f
|
|
258
|
+
fin
|
|
259
|
+
final
|
|
260
|
+
fue
|
|
261
|
+
fuera
|
|
262
|
+
fuerais
|
|
263
|
+
fueran
|
|
264
|
+
fueras
|
|
265
|
+
fueron
|
|
266
|
+
fuese
|
|
267
|
+
fueseis
|
|
268
|
+
fuesen
|
|
269
|
+
fueses
|
|
270
|
+
fui
|
|
271
|
+
fuimos
|
|
272
|
+
fuiste
|
|
273
|
+
fuisteis
|
|
274
|
+
fuéramos
|
|
275
|
+
fuésemos
|
|
276
|
+
g
|
|
277
|
+
general
|
|
278
|
+
gran
|
|
279
|
+
grandes
|
|
280
|
+
gueno
|
|
281
|
+
h
|
|
282
|
+
ha
|
|
283
|
+
haber
|
|
284
|
+
habia
|
|
285
|
+
habida
|
|
286
|
+
habidas
|
|
287
|
+
habido
|
|
288
|
+
habidos
|
|
289
|
+
habiendo
|
|
290
|
+
habla
|
|
291
|
+
hablan
|
|
292
|
+
habremos
|
|
293
|
+
habrá
|
|
294
|
+
habrán
|
|
295
|
+
habrás
|
|
296
|
+
habré
|
|
297
|
+
habréis
|
|
298
|
+
habría
|
|
299
|
+
habríais
|
|
300
|
+
habríamos
|
|
301
|
+
habrían
|
|
302
|
+
habrías
|
|
303
|
+
habéis
|
|
304
|
+
había
|
|
305
|
+
habíais
|
|
306
|
+
habíamos
|
|
307
|
+
habían
|
|
308
|
+
habías
|
|
309
|
+
hace
|
|
310
|
+
haceis
|
|
311
|
+
hacemos
|
|
312
|
+
hacen
|
|
313
|
+
hacer
|
|
314
|
+
hacerlo
|
|
315
|
+
haces
|
|
316
|
+
hacia
|
|
317
|
+
haciendo
|
|
318
|
+
hago
|
|
319
|
+
han
|
|
320
|
+
has
|
|
321
|
+
hasta
|
|
322
|
+
hay
|
|
323
|
+
haya
|
|
324
|
+
hayamos
|
|
325
|
+
hayan
|
|
326
|
+
hayas
|
|
327
|
+
hayáis
|
|
328
|
+
he
|
|
329
|
+
hecho
|
|
330
|
+
hemos
|
|
331
|
+
hicieron
|
|
332
|
+
hizo
|
|
333
|
+
horas
|
|
334
|
+
hoy
|
|
335
|
+
hube
|
|
336
|
+
hubiera
|
|
337
|
+
hubierais
|
|
338
|
+
hubieran
|
|
339
|
+
hubieras
|
|
340
|
+
hubieron
|
|
341
|
+
hubiese
|
|
342
|
+
hubieseis
|
|
343
|
+
hubiesen
|
|
344
|
+
hubieses
|
|
345
|
+
hubimos
|
|
346
|
+
hubiste
|
|
347
|
+
hubisteis
|
|
348
|
+
hubiéramos
|
|
349
|
+
hubiésemos
|
|
350
|
+
hubo
|
|
351
|
+
i
|
|
352
|
+
igual
|
|
353
|
+
incluso
|
|
354
|
+
indicó
|
|
355
|
+
informo
|
|
356
|
+
informó
|
|
357
|
+
intenta
|
|
358
|
+
intentais
|
|
359
|
+
intentamos
|
|
360
|
+
intentan
|
|
361
|
+
intentar
|
|
362
|
+
intentas
|
|
363
|
+
intento
|
|
364
|
+
ir
|
|
365
|
+
j
|
|
366
|
+
junto
|
|
367
|
+
k
|
|
368
|
+
l
|
|
369
|
+
la
|
|
370
|
+
lado
|
|
371
|
+
largo
|
|
372
|
+
las
|
|
373
|
+
le
|
|
374
|
+
lejos
|
|
375
|
+
les
|
|
376
|
+
llegó
|
|
377
|
+
lleva
|
|
378
|
+
llevar
|
|
379
|
+
lo
|
|
380
|
+
los
|
|
381
|
+
luego
|
|
382
|
+
lugar
|
|
383
|
+
m
|
|
384
|
+
mal
|
|
385
|
+
manera
|
|
386
|
+
manifestó
|
|
387
|
+
mas
|
|
388
|
+
mayor
|
|
389
|
+
me
|
|
390
|
+
mediante
|
|
391
|
+
medio
|
|
392
|
+
mejor
|
|
393
|
+
mencionó
|
|
394
|
+
menos
|
|
395
|
+
menudo
|
|
396
|
+
mi
|
|
397
|
+
mia
|
|
398
|
+
mias
|
|
399
|
+
mientras
|
|
400
|
+
mio
|
|
401
|
+
mios
|
|
402
|
+
mis
|
|
403
|
+
misma
|
|
404
|
+
mismas
|
|
405
|
+
mismo
|
|
406
|
+
mismos
|
|
407
|
+
modo
|
|
408
|
+
momento
|
|
409
|
+
mucha
|
|
410
|
+
muchas
|
|
411
|
+
mucho
|
|
412
|
+
muchos
|
|
413
|
+
muy
|
|
414
|
+
más
|
|
415
|
+
mí
|
|
416
|
+
mía
|
|
417
|
+
mías
|
|
418
|
+
mío
|
|
419
|
+
míos
|
|
420
|
+
n
|
|
421
|
+
nada
|
|
422
|
+
nadie
|
|
423
|
+
ni
|
|
424
|
+
ninguna
|
|
425
|
+
ningunas
|
|
426
|
+
ninguno
|
|
427
|
+
ningunos
|
|
428
|
+
ningún
|
|
429
|
+
no
|
|
430
|
+
nos
|
|
431
|
+
nosotras
|
|
432
|
+
nosotros
|
|
433
|
+
nuestra
|
|
434
|
+
nuestras
|
|
435
|
+
nuestro
|
|
436
|
+
nuestros
|
|
437
|
+
nueva
|
|
438
|
+
nuevas
|
|
439
|
+
nuevo
|
|
440
|
+
nuevos
|
|
441
|
+
nunca
|
|
442
|
+
o
|
|
443
|
+
ocho
|
|
444
|
+
os
|
|
445
|
+
otra
|
|
446
|
+
otras
|
|
447
|
+
otro
|
|
448
|
+
otros
|
|
449
|
+
p
|
|
450
|
+
pais
|
|
451
|
+
para
|
|
452
|
+
parece
|
|
453
|
+
parte
|
|
454
|
+
partir
|
|
455
|
+
pasada
|
|
456
|
+
pasado
|
|
457
|
+
paìs
|
|
458
|
+
peor
|
|
459
|
+
pero
|
|
460
|
+
pesar
|
|
461
|
+
poca
|
|
462
|
+
pocas
|
|
463
|
+
poco
|
|
464
|
+
pocos
|
|
465
|
+
podeis
|
|
466
|
+
podemos
|
|
467
|
+
poder
|
|
468
|
+
podria
|
|
469
|
+
podriais
|
|
470
|
+
podriamos
|
|
471
|
+
podrian
|
|
472
|
+
podrias
|
|
473
|
+
podrá
|
|
474
|
+
podrán
|
|
475
|
+
podría
|
|
476
|
+
podrían
|
|
477
|
+
poner
|
|
478
|
+
por
|
|
479
|
+
por qué
|
|
480
|
+
porque
|
|
481
|
+
posible
|
|
482
|
+
primer
|
|
483
|
+
primera
|
|
484
|
+
primero
|
|
485
|
+
primeros
|
|
486
|
+
principalmente
|
|
487
|
+
pronto
|
|
488
|
+
propia
|
|
489
|
+
propias
|
|
490
|
+
propio
|
|
491
|
+
propios
|
|
492
|
+
proximo
|
|
493
|
+
próximo
|
|
494
|
+
próximos
|
|
495
|
+
pudo
|
|
496
|
+
pueda
|
|
497
|
+
puede
|
|
498
|
+
pueden
|
|
499
|
+
puedo
|
|
500
|
+
pues
|
|
501
|
+
q
|
|
502
|
+
qeu
|
|
503
|
+
que
|
|
504
|
+
quedó
|
|
505
|
+
queremos
|
|
506
|
+
quien
|
|
507
|
+
quienes
|
|
508
|
+
quiere
|
|
509
|
+
quiza
|
|
510
|
+
quizas
|
|
511
|
+
quizá
|
|
512
|
+
quizás
|
|
513
|
+
quién
|
|
514
|
+
quiénes
|
|
515
|
+
qué
|
|
516
|
+
r
|
|
517
|
+
raras
|
|
518
|
+
realizado
|
|
519
|
+
realizar
|
|
520
|
+
realizó
|
|
521
|
+
repente
|
|
522
|
+
respecto
|
|
523
|
+
s
|
|
524
|
+
sabe
|
|
525
|
+
sabeis
|
|
526
|
+
sabemos
|
|
527
|
+
saben
|
|
528
|
+
saber
|
|
529
|
+
sabes
|
|
530
|
+
sal
|
|
531
|
+
salvo
|
|
532
|
+
se
|
|
533
|
+
sea
|
|
534
|
+
seamos
|
|
535
|
+
sean
|
|
536
|
+
seas
|
|
537
|
+
segun
|
|
538
|
+
segunda
|
|
539
|
+
segundo
|
|
540
|
+
según
|
|
541
|
+
seis
|
|
542
|
+
ser
|
|
543
|
+
sera
|
|
544
|
+
seremos
|
|
545
|
+
será
|
|
546
|
+
serán
|
|
547
|
+
serás
|
|
548
|
+
seré
|
|
549
|
+
seréis
|
|
550
|
+
sería
|
|
551
|
+
seríais
|
|
552
|
+
seríamos
|
|
553
|
+
serían
|
|
554
|
+
serías
|
|
555
|
+
seáis
|
|
556
|
+
señaló
|
|
557
|
+
si
|
|
558
|
+
sido
|
|
559
|
+
siempre
|
|
560
|
+
siendo
|
|
561
|
+
siete
|
|
562
|
+
sigue
|
|
563
|
+
siguiente
|
|
564
|
+
sin
|
|
565
|
+
sino
|
|
566
|
+
sobre
|
|
567
|
+
sois
|
|
568
|
+
sola
|
|
569
|
+
solamente
|
|
570
|
+
solas
|
|
571
|
+
solo
|
|
572
|
+
solos
|
|
573
|
+
somos
|
|
574
|
+
son
|
|
575
|
+
soy
|
|
576
|
+
soyos
|
|
577
|
+
su
|
|
578
|
+
supuesto
|
|
579
|
+
sus
|
|
580
|
+
suya
|
|
581
|
+
suyas
|
|
582
|
+
suyo
|
|
583
|
+
suyos
|
|
584
|
+
sé
|
|
585
|
+
sí
|
|
586
|
+
sólo
|
|
587
|
+
t
|
|
588
|
+
tal
|
|
589
|
+
tambien
|
|
590
|
+
también
|
|
591
|
+
tampoco
|
|
592
|
+
tan
|
|
593
|
+
tanto
|
|
594
|
+
tarde
|
|
595
|
+
te
|
|
596
|
+
temprano
|
|
597
|
+
tendremos
|
|
598
|
+
tendrá
|
|
599
|
+
tendrán
|
|
600
|
+
tendrás
|
|
601
|
+
tendré
|
|
602
|
+
tendréis
|
|
603
|
+
tendría
|
|
604
|
+
tendríais
|
|
605
|
+
tendríamos
|
|
606
|
+
tendrían
|
|
607
|
+
tendrías
|
|
608
|
+
tened
|
|
609
|
+
teneis
|
|
610
|
+
tenemos
|
|
611
|
+
tener
|
|
612
|
+
tenga
|
|
613
|
+
tengamos
|
|
614
|
+
tengan
|
|
615
|
+
tengas
|
|
616
|
+
tengo
|
|
617
|
+
tengáis
|
|
618
|
+
tenida
|
|
619
|
+
tenidas
|
|
620
|
+
tenido
|
|
621
|
+
tenidos
|
|
622
|
+
teniendo
|
|
623
|
+
tenéis
|
|
624
|
+
tenía
|
|
625
|
+
teníais
|
|
626
|
+
teníamos
|
|
627
|
+
tenían
|
|
628
|
+
tenías
|
|
629
|
+
tercera
|
|
630
|
+
ti
|
|
631
|
+
tiempo
|
|
632
|
+
tiene
|
|
633
|
+
tienen
|
|
634
|
+
tienes
|
|
635
|
+
toda
|
|
636
|
+
todas
|
|
637
|
+
todavia
|
|
638
|
+
todavía
|
|
639
|
+
todo
|
|
640
|
+
todos
|
|
641
|
+
total
|
|
642
|
+
trabaja
|
|
643
|
+
trabajais
|
|
644
|
+
trabajamos
|
|
645
|
+
trabajan
|
|
646
|
+
trabajar
|
|
647
|
+
trabajas
|
|
648
|
+
trabajo
|
|
649
|
+
tras
|
|
650
|
+
trata
|
|
651
|
+
través
|
|
652
|
+
tres
|
|
653
|
+
tu
|
|
654
|
+
tus
|
|
655
|
+
tuve
|
|
656
|
+
tuviera
|
|
657
|
+
tuvierais
|
|
658
|
+
tuvieran
|
|
659
|
+
tuvieras
|
|
660
|
+
tuvieron
|
|
661
|
+
tuviese
|
|
662
|
+
tuvieseis
|
|
663
|
+
tuviesen
|
|
664
|
+
tuvieses
|
|
665
|
+
tuvimos
|
|
666
|
+
tuviste
|
|
667
|
+
tuvisteis
|
|
668
|
+
tuviéramos
|
|
669
|
+
tuviésemos
|
|
670
|
+
tuvo
|
|
671
|
+
tuya
|
|
672
|
+
tuyas
|
|
673
|
+
tuyo
|
|
674
|
+
tuyos
|
|
675
|
+
tú
|
|
676
|
+
u
|
|
677
|
+
ultimo
|
|
678
|
+
un
|
|
679
|
+
una
|
|
680
|
+
unas
|
|
681
|
+
uno
|
|
682
|
+
unos
|
|
683
|
+
usa
|
|
684
|
+
usais
|
|
685
|
+
usamos
|
|
686
|
+
usan
|
|
687
|
+
usar
|
|
688
|
+
usas
|
|
689
|
+
uso
|
|
690
|
+
usted
|
|
691
|
+
ustedes
|
|
692
|
+
v
|
|
693
|
+
va
|
|
694
|
+
vais
|
|
695
|
+
valor
|
|
696
|
+
vamos
|
|
697
|
+
van
|
|
698
|
+
varias
|
|
699
|
+
varios
|
|
700
|
+
vaya
|
|
701
|
+
veces
|
|
702
|
+
ver
|
|
703
|
+
verdad
|
|
704
|
+
verdadera
|
|
705
|
+
verdadero
|
|
706
|
+
vez
|
|
707
|
+
vosotras
|
|
708
|
+
vosotros
|
|
709
|
+
voy
|
|
710
|
+
vuestra
|
|
711
|
+
vuestras
|
|
712
|
+
vuestro
|
|
713
|
+
vuestros
|
|
714
|
+
w
|
|
715
|
+
x
|
|
716
|
+
y
|
|
717
|
+
ya
|
|
718
|
+
yo
|
|
719
|
+
z
|
|
720
|
+
él
|
|
721
|
+
éramos
|
|
722
|
+
ésa
|
|
723
|
+
ésas
|
|
724
|
+
ése
|
|
725
|
+
ésos
|
|
726
|
+
ésta
|
|
727
|
+
éstas
|
|
728
|
+
éste
|
|
729
|
+
éstos
|
|
730
|
+
última
|
|
731
|
+
últimas
|
|
732
|
+
último
|
|
733
|
+
últimos
|
data/lib/raske-nlp.rb
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
1
|
+
require "raske-nlp/version"
|
|
2
|
+
require "raske-nlp/stop_list"
|
|
3
|
+
require "raske-nlp/result"
|
|
4
|
+
|
|
5
|
+
module RaskeNLP
|
|
6
|
+
DEFAULTS = {
|
|
7
|
+
min_phrase_length: 1,
|
|
8
|
+
max_phrase_length: 3,
|
|
9
|
+
min_frequency: 1,
|
|
10
|
+
min_score: 1,
|
|
11
|
+
stop_list: StopList::ISO
|
|
12
|
+
}
|
|
13
|
+
|
|
14
|
+
def self.run(text, options={})
|
|
15
|
+
RaskeNLP::Result.new(text, DEFAULTS.merge(options))
|
|
16
|
+
end
|
|
17
|
+
end
|
|
@@ -0,0 +1,79 @@
|
|
|
1
|
+
module RaskeNLP
|
|
2
|
+
class Result
|
|
3
|
+
SENTENCE_REGEX = /[\\[\\]\n.!?,;:\t\\-\\"\\(\\)\\\'\u2019\u2013]/
|
|
4
|
+
WORD_REGEX = /[^a-zA-Z0-9_\+\-\/]/
|
|
5
|
+
NUMBER_REGEX = /^-*[0-9,\.]+$/
|
|
6
|
+
|
|
7
|
+
def initialize(text, options={})
|
|
8
|
+
@text, @options = text, options
|
|
9
|
+
@stoplist_regex = options[:stop_list].to_regex
|
|
10
|
+
end
|
|
11
|
+
|
|
12
|
+
def sentences
|
|
13
|
+
@text.split(SENTENCE_REGEX).map(&:strip)
|
|
14
|
+
end
|
|
15
|
+
|
|
16
|
+
def phrases
|
|
17
|
+
@phrases ||= begin
|
|
18
|
+
sentences.map { |sentence|
|
|
19
|
+
sentence.downcase.split(@stoplist_regex).map(&:strip).select { |phrase| acceptable?(phrase) }
|
|
20
|
+
}.flatten
|
|
21
|
+
end
|
|
22
|
+
end
|
|
23
|
+
|
|
24
|
+
def word_scores
|
|
25
|
+
@word_scores ||= begin
|
|
26
|
+
frequencies = Hash.new(0)
|
|
27
|
+
degrees = Hash.new(0)
|
|
28
|
+
|
|
29
|
+
phrases.each do |phrase|
|
|
30
|
+
words = split_words(phrase)
|
|
31
|
+
words.each do |word|
|
|
32
|
+
frequencies[word] += 1
|
|
33
|
+
degrees[word] += words.length - 1
|
|
34
|
+
end
|
|
35
|
+
end
|
|
36
|
+
|
|
37
|
+
frequencies.each do |word, frequency|
|
|
38
|
+
degrees[word] += frequency
|
|
39
|
+
end
|
|
40
|
+
|
|
41
|
+
scores = {}
|
|
42
|
+
frequencies.each do |word, frequency|
|
|
43
|
+
scores[word] = degrees[word] / frequency.to_f
|
|
44
|
+
end
|
|
45
|
+
scores
|
|
46
|
+
end
|
|
47
|
+
end
|
|
48
|
+
|
|
49
|
+
def keywords
|
|
50
|
+
keywords = Hash.new(0)
|
|
51
|
+
|
|
52
|
+
phrases.each do |phrase|
|
|
53
|
+
if @options[:min_frequency] > 1
|
|
54
|
+
next if phrases.count(phrase) < @options[:min_frequency]
|
|
55
|
+
end
|
|
56
|
+
|
|
57
|
+
words = split_words(phrase)
|
|
58
|
+
keywords[phrase] = words.map { |word| word_scores[word] }.sum
|
|
59
|
+
end
|
|
60
|
+
|
|
61
|
+
keywords.select { |word, score|
|
|
62
|
+
score >= @options[:min_score]
|
|
63
|
+
}.sort_by(&:last).reverse.to_h
|
|
64
|
+
end
|
|
65
|
+
|
|
66
|
+
private
|
|
67
|
+
def split_words(str)
|
|
68
|
+
str.split(WORD_REGEX).reject { |word| word =~ NUMBER_REGEX }
|
|
69
|
+
end
|
|
70
|
+
|
|
71
|
+
def acceptable?(phrase)
|
|
72
|
+
!phrase.empty? &&
|
|
73
|
+
phrase.length >= @options[:min_phrase_length] &&
|
|
74
|
+
phrase.split(/\s/).length <= @options[:max_phrase_length] &&
|
|
75
|
+
phrase.scan(/[a-zA-Z]/).length > 0 &&
|
|
76
|
+
phrase.scan(/[a-zA-Z]/).length > phrase.scan(/\d/).length
|
|
77
|
+
end
|
|
78
|
+
end
|
|
79
|
+
end
|
|
@@ -0,0 +1,22 @@
|
|
|
1
|
+
module RaskeNLP
|
|
2
|
+
class StopList
|
|
3
|
+
def initialize(words)
|
|
4
|
+
@words = words
|
|
5
|
+
end
|
|
6
|
+
|
|
7
|
+
def to_a
|
|
8
|
+
@words
|
|
9
|
+
end
|
|
10
|
+
|
|
11
|
+
def to_regex
|
|
12
|
+
@regex ||= /#{@words.map { |w| "\\b#{w}(?![\\w-])" }.join("|")}/i
|
|
13
|
+
end
|
|
14
|
+
|
|
15
|
+
def self.read(filename)
|
|
16
|
+
new(File.read(filename).split("\n").reject { |line| line =~ /^#/ })
|
|
17
|
+
end
|
|
18
|
+
|
|
19
|
+
ISO = read(File.join(File.dirname(__FILE__), "../../data/StopWordsIso.txt"))
|
|
20
|
+
ALI = read(File.join(File.dirname(__FILE__), "../../data/AliStopList.txt"))
|
|
21
|
+
end
|
|
22
|
+
end
|
data/raske-nlp.gemspec
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
1
|
+
lib = File.expand_path("../lib", __FILE__)
|
|
2
|
+
$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
|
|
3
|
+
require "raske-nlp/version"
|
|
4
|
+
|
|
5
|
+
Gem::Specification.new do |spec|
|
|
6
|
+
spec.name = "raske-nlp"
|
|
7
|
+
spec.version = RaskeNLP::VERSION
|
|
8
|
+
spec.authors = ["Adrián Pradilla Pórtoles"]
|
|
9
|
+
spec.email = ["adrianpradilla@gmail.com"]
|
|
10
|
+
|
|
11
|
+
spec.summary = "Implementation of the Rapid Automatic Spanish Keyword Extraction (RAKE) algorithm"
|
|
12
|
+
spec.description = "Implementation of the Rapid Automatic Spanish Keyword Extraction (RAKE) algorithm"
|
|
13
|
+
|
|
14
|
+
spec.files = `git ls-files -z`.split("\x0").reject do |f|
|
|
15
|
+
f.match(%r{^(test|spec|features)/})
|
|
16
|
+
end
|
|
17
|
+
spec.bindir = "exe"
|
|
18
|
+
spec.executables = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
|
|
19
|
+
spec.require_paths = ["lib"]
|
|
20
|
+
|
|
21
|
+
spec.add_development_dependency "bundler", "~> 1.16"
|
|
22
|
+
spec.add_development_dependency "rake", "~> 13.0"
|
|
23
|
+
end
|
metadata
ADDED
|
@@ -0,0 +1,86 @@
|
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
|
2
|
+
name: raske-nlp
|
|
3
|
+
version: !ruby/object:Gem::Version
|
|
4
|
+
version: 0.1.5
|
|
5
|
+
platform: ruby
|
|
6
|
+
authors:
|
|
7
|
+
- Adrián Pradilla Pórtoles
|
|
8
|
+
autorequire:
|
|
9
|
+
bindir: exe
|
|
10
|
+
cert_chain: []
|
|
11
|
+
date: 2021-04-10 00:00:00.000000000 Z
|
|
12
|
+
dependencies:
|
|
13
|
+
- !ruby/object:Gem::Dependency
|
|
14
|
+
name: bundler
|
|
15
|
+
requirement: !ruby/object:Gem::Requirement
|
|
16
|
+
requirements:
|
|
17
|
+
- - "~>"
|
|
18
|
+
- !ruby/object:Gem::Version
|
|
19
|
+
version: '1.16'
|
|
20
|
+
type: :development
|
|
21
|
+
prerelease: false
|
|
22
|
+
version_requirements: !ruby/object:Gem::Requirement
|
|
23
|
+
requirements:
|
|
24
|
+
- - "~>"
|
|
25
|
+
- !ruby/object:Gem::Version
|
|
26
|
+
version: '1.16'
|
|
27
|
+
- !ruby/object:Gem::Dependency
|
|
28
|
+
name: rake
|
|
29
|
+
requirement: !ruby/object:Gem::Requirement
|
|
30
|
+
requirements:
|
|
31
|
+
- - "~>"
|
|
32
|
+
- !ruby/object:Gem::Version
|
|
33
|
+
version: '13.0'
|
|
34
|
+
type: :development
|
|
35
|
+
prerelease: false
|
|
36
|
+
version_requirements: !ruby/object:Gem::Requirement
|
|
37
|
+
requirements:
|
|
38
|
+
- - "~>"
|
|
39
|
+
- !ruby/object:Gem::Version
|
|
40
|
+
version: '13.0'
|
|
41
|
+
description: Implementation of the Rapid Automatic Spanish Keyword Extraction (RAKE)
|
|
42
|
+
algorithm
|
|
43
|
+
email:
|
|
44
|
+
- adrianpradilla@gmail.com
|
|
45
|
+
executables: []
|
|
46
|
+
extensions: []
|
|
47
|
+
extra_rdoc_files: []
|
|
48
|
+
files:
|
|
49
|
+
- ".gitignore"
|
|
50
|
+
- ".travis.yml"
|
|
51
|
+
- Gemfile
|
|
52
|
+
- Gemfile.lock
|
|
53
|
+
- MIT-LICENSE
|
|
54
|
+
- README.md
|
|
55
|
+
- Rakefile
|
|
56
|
+
- bin/console
|
|
57
|
+
- data/AliStopList.txt
|
|
58
|
+
- data/StopWordsIso.txt
|
|
59
|
+
- lib/raske-nlp.rb
|
|
60
|
+
- lib/raske-nlp/result.rb
|
|
61
|
+
- lib/raske-nlp/stop_list.rb
|
|
62
|
+
- lib/raske-nlp/version.rb
|
|
63
|
+
- raske-nlp.gemspec
|
|
64
|
+
homepage:
|
|
65
|
+
licenses: []
|
|
66
|
+
metadata: {}
|
|
67
|
+
post_install_message:
|
|
68
|
+
rdoc_options: []
|
|
69
|
+
require_paths:
|
|
70
|
+
- lib
|
|
71
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
|
72
|
+
requirements:
|
|
73
|
+
- - ">="
|
|
74
|
+
- !ruby/object:Gem::Version
|
|
75
|
+
version: '0'
|
|
76
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
77
|
+
requirements:
|
|
78
|
+
- - ">="
|
|
79
|
+
- !ruby/object:Gem::Version
|
|
80
|
+
version: '0'
|
|
81
|
+
requirements: []
|
|
82
|
+
rubygems_version: 3.0.8
|
|
83
|
+
signing_key:
|
|
84
|
+
specification_version: 4
|
|
85
|
+
summary: Implementation of the Rapid Automatic Spanish Keyword Extraction (RAKE) algorithm
|
|
86
|
+
test_files: []
|