classifier-reborn 2.1.0 → 2.2.0

Sign up to get free protection for your applications and to get access to all the features.
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA1:
3
- metadata.gz: 53745cead2833e74e9d74f3359bae0bc7fd01fa4
4
- data.tar.gz: 2926365890cf0bc43f7ff17570789ed5e032a45d
3
+ metadata.gz: e626667dbd70c34cda4604852500af1bd9cc8f9f
4
+ data.tar.gz: 36dad105dca4770c1a5b708257c66721522d235e
5
5
  SHA512:
6
- metadata.gz: 5def462eccbb9ef7a45d3968a5fbc2e2a886bfdda734a98e1f5904943ba1b4012ea331a51ff52e2c165db776dc8c6aed35fac9e01ea3e3f001f82daf98027f74
7
- data.tar.gz: 22726b8f6c2acab5bb47b9faac2a2d615557bfff535327ef5bbe022a661c1be13c3b7cec95d1c0030f93158eb0942e121d97f2a0213a640eca00607de9d55f61
6
+ metadata.gz: aa08b0c7ad09138ac9efb0f55daa3d24a43b1c1388cef4ad2a71a9375620ce1794db66d3eca470c7772a80e4eac8e3c7a6811b20a3b65e2e93cfced86922554b
7
+ data.tar.gz: c137296ec3661043828e99abd11814b44a1d7e7ab6ca3d6f36ae54eaba1044d426847a9782bd106e082a6f7bacc22d72910eb51e6a37516017886a4c7df38a57
data/LICENSE CHANGED
@@ -426,4 +426,77 @@ the Free Software Foundation.
426
426
  14. If you wish to incorporate parts of the Library into other free
427
427
  programs whose distribution conditions are incompatible with these,
428
428
  write to the author to ask for permission. For software which is
429
- copyrighted by
429
+ copyrighted by the Free Software Foundation, write to the Free
430
+ Software Foundation; we sometimes make exceptions for this. Our
431
+ decision will be guided by the two goals of preserving the free status
432
+ of all derivatives of our free software and of promoting the sharing
433
+ and reuse of software generally.
434
+
435
+ NO WARRANTY
436
+
437
+ 15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
438
+ WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
439
+ EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
440
+ OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY
441
+ KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
442
+ IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
443
+ PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
444
+ LIBRARY IS WITH YOU. SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME
445
+ THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
446
+
447
+ 16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
448
+ WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY
449
+ AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU
450
+ FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
451
+ CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE
452
+ LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
453
+ RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A
454
+ FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
455
+ SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
456
+ DAMAGES.
457
+
458
+ END OF TERMS AND CONDITIONS
459
+
460
+ How to Apply These Terms to Your New Libraries
461
+
462
+ If you develop a new library, and you want it to be of the greatest
463
+ possible use to the public, we recommend making it free software that
464
+ everyone can redistribute and change. You can do so by permitting
465
+ redistribution under these terms (or, alternatively, under the terms of the
466
+ ordinary General Public License).
467
+
468
+ To apply these terms, attach the following notices to the library. It is
469
+ safest to attach them to the start of each source file to most effectively
470
+ convey the exclusion of warranty; and each file should have at least the
471
+ "copyright" line and a pointer to where the full notice is found.
472
+
473
+ <one line to give the library's name and a brief idea of what it does.>
474
+ Copyright (C) <year> <name of author>
475
+
476
+ This library is free software; you can redistribute it and/or
477
+ modify it under the terms of the GNU Lesser General Public
478
+ License as published by the Free Software Foundation; either
479
+ version 2.1 of the License, or (at your option) any later version.
480
+
481
+ This library is distributed in the hope that it will be useful,
482
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
483
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
484
+ Lesser General Public License for more details.
485
+
486
+ You should have received a copy of the GNU Lesser General Public
487
+ License along with this library; if not, write to the Free Software
488
+ Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
489
+
490
+ Also add information on how to contact you by electronic and paper mail.
491
+
492
+ You should also get your employer (if you work as a programmer) or your
493
+ school, if any, to sign a "copyright disclaimer" for the library, if
494
+ necessary. Here is a sample; alter the names:
495
+
496
+ Yoyodyne, Inc., hereby disclaims all copyright interest in the
497
+ library `Frob' (a library for tweaking knobs) written by James Random Hacker.
498
+
499
+ <signature of Ty Coon>, 1 April 1990
500
+ Ty Coon, President of Vice
501
+
502
+ That's all there is to it!
@@ -1,252 +1,82 @@
1
- ## Welcome to Classifier Reborn
1
+ # Classifier Reborn
2
2
 
3
- [![Gem Version](https://img.shields.io/gem/v/classifier-reborn.svg)][ruby-gems]
4
- [![Build Status](https://img.shields.io/travis/jekyll/classifier-reborn/master.svg)][travis]
5
- [![Dependency Status](https://img.shields.io/gemnasium/jekyll/classifier-reborn.svg)][gemnasium]
6
- [ruby-gems]: https://rubygems.org/gems/jekyll/classifier-reborn
7
- [gemnasium]: https://gemnasium.com/jekyll/classifier-reborn
8
- [travis]: https://travis-ci.org/jekyll/classifier-reborn
3
+ [![Gem Version](https://badge.fury.io/rb/classifier-reborn.svg)](https://rubygems.org/gems/classifier-reborn)
4
+ [![Build Status](https://img.shields.io/travis/jekyll/classifier-reborn/master.svg)](https://travis-ci.org/jekyll/classifier-reborn)
5
+ [![Dependency Status](https://img.shields.io/gemnasium/jekyll/classifier-reborn.svg)](https://gemnasium.com/jekyll/classifier-reborn)
9
6
 
10
- Classifier is a general module to allow Bayesian and other types of classifications.
7
+ ---
11
8
 
12
- Classifier Reborn is a fork of cardmagic/classifier under more active development.
9
+ ## [Read the Docs](http://www.classifier-reborn.com/)
13
10
 
14
- ## Download
11
+ ## Getting Started
15
12
 
16
- Add this line to your application's Gemfile:
13
+ Classifier Reborn is a general classifier module to allow Bayesian and other types of classifications.
14
+ It is a fork of [cardmagic/classifier](https://github.com/cardmagic/classifier) under more active development.
15
+ Currently, it has [Bayesian Classifier](https://en.wikipedia.org/wiki/Naive_Bayes_classifier) and [Latent Semantic Indexer (LSI)](https://en.wikipedia.org/wiki/Latent_semantic_analysis) implemented.
17
16
 
18
- gem 'classifier-reborn'
17
+ Here is a quick illustration of the Bayesian classifier.
19
18
 
20
- And then execute:
21
-
22
- $ bundle
23
-
24
- Or install it yourself as:
25
-
26
- $ gem install classifier-reborn
27
-
28
- ## Dependencies
29
-
30
- The only runtime dependency you'll need to install is Roman Shterenzon's fast-stemmer gem:
31
-
32
- gem install fast-stemmer
33
-
34
- This should install automatically with RubyGems.
35
-
36
- If you would like to speed up LSI classification by at least 10x, please install the following libraries:
37
-
38
- * [GNU GSL](http://www.gnu.org/software/gsl)
39
- * [rb-gsl](https://rubygems.org/gems/rb-gsl)
40
-
41
- Notice that LSI will work without these libraries, but as soon as they are installed, Classifier will make use of them. No configuration changes are needed, we like to keep things ridiculously easy for you.
42
-
43
- ## Bayes
44
-
45
- A Bayesian classifier by Lucas Carlson. Bayesian Classifiers are accurate, fast, and have modest memory requirements.
46
-
47
- *Note: Classifier only supports UTF-8 characters.*
48
-
49
- ### Usage
50
-
51
- ```ruby
52
- require 'classifier-reborn'
53
- classifier = ClassifierReborn::Bayes.new 'Interesting', 'Uninteresting'
54
- classifier.train_interesting "here are some good words. I hope you love them"
55
- classifier.train_uninteresting "here are some bad words, I hate you"
56
- classifier.classify "I hate bad words and you" # returns 'Uninteresting'
57
-
58
- classifier_snapshot = Marshal.dump classifier
59
- # This is a string of bytes, you can persist it anywhere you like
60
-
61
- File.open("classifier.dat", "w") {|f| f.write(classifier_snapshot) }
62
- # Or Redis.current.save "classifier", classifier_snapshot
63
-
64
- # This is now saved to a file, and you can safely restart the application
65
- data = File.read("classifier.dat")
66
- # Or data = Redis.current.get "classifier"
67
- trained_classifier = Marshal.load data
68
- trained_classifier.classify "I love" # returns 'Interesting'
19
+ ```bash
20
+ $ gem install classifier-reborn
21
+ $ irb
22
+ irb(main):001:0> require 'classifier-reborn'
23
+ irb(main):002:0> classifier = ClassifierReborn::Bayes.new 'Ham', 'Spam'
24
+ irb(main):003:0> classifier.train "Ham", "Sunday is a holiday. Say no to work on Sunday!"
25
+ irb(main):004:0> classifier.train "Spam", "You are the lucky winner! Claim your holiday prize."
26
+ irb(main):005:0> classifier.classify "What's the plan for Sunday?"
27
+ #=> "Ham"
69
28
  ```
70
29
 
71
- Beyond the basic example, the constructor and trainer can be used in a more
72
- flexible way to accommodate non-trival applications. Consider the following
73
- program:
74
-
75
- ```ruby
76
- #!/usr/bin/env ruby
77
- # classifier_reborn_demo.rb
78
-
79
- require 'classifier-reborn'
80
-
81
- training_set = DATA.read.split("\n")
82
- categories = training_set.shift.split(',').map{|c| c.strip}
83
-
84
- # pass :auto_categorize option to allow feeding previously unknown categories
85
- classifier = ClassifierReborn::Bayes.new categories, auto_categorize: true
86
-
87
- training_set.each do |a_line|
88
- next if a_line.empty? || '#' == a_line.strip[0]
89
- parts = a_line.strip.split(':')
90
- classifier.train(parts.first, parts.last)
91
- end
92
-
93
- puts classifier.classify "I hate bad words and you" #=> 'Uninteresting'
94
- puts classifier.classify "I hate javascript" #=> 'Uninteresting'
95
- puts classifier.classify "javascript is bad" #=> 'Uninteresting'
96
-
97
- puts classifier.classify "all you need is ruby" #=> 'Interesting'
98
- puts classifier.classify "i love ruby" #=> 'Interesting'
99
-
100
- puts classifier.classify "which is better dogs or cats" #=> 'dog'
101
- puts classifier.classify "what do I need to kill rats and mice" #=> 'cat'
102
-
103
- __END__
104
- Interesting, Uninteresting
105
- interesting: here are some good words. I hope you love them
106
- interesting: all you need is love
107
- interesting: the love boat, soon we will be taking another ride
108
- interesting: ruby don't take your love to town
109
-
110
- uninteresting: here are some bad words, I hate you
111
- uninteresting: bad bad leroy brown badest man in the darn town
112
- uninteresting: the good the bad and the ugly
113
- uninteresting: java, javascript, css front-end html
114
- #
115
- # train categories that were not pre-described
116
- #
117
- dog: dog days of summer
118
- dog: a man's best friend is his dog
119
- dog: a good hunting dog is a fine thing
120
- dog: man my dogs are tired
121
- dog: dogs are better than cats in soooo many ways
122
-
123
- cat: the fuzz ball spilt the milk
124
- cat: got rats or mice get a cat to kill them
125
- cat: cats never come when you call them
126
- cat: That dang cat keeps scratching the furniture
30
+ Now, let's build an LSI, classify some text, and find a cluster of related documents.
31
+
32
+ ```bash
33
+ irb(main):006:0> lsi = ClassifierReborn::LSI.new
34
+ irb(main):007:0> lsi.add_item "This text deals with dogs. Dogs.", :dog
35
+ irb(main):008:0> lsi.add_item "This text involves dogs too. Dogs!", :dog
36
+ irb(main):009:0> lsi.add_item "This text revolves around cats. Cats.", :cat
37
+ irb(main):010:0> lsi.add_item "This text also involves cats. Cats!", :cat
38
+ irb(main):011:0> lsi.add_item "This text involves birds. Birds.", :bird
39
+ irb(main):012:0> lsi.classify "This text is about dogs!"
40
+ #=> :dog
41
+ irb(main):013:0> lsi.find_related("This text is around cats!", 2)
42
+ #=> ["This text revolves around cats. Cats.", "This text also involves cats. Cats!"]
127
43
  ```
128
44
 
129
- #### Knowing the Score
130
-
131
- When you ask a bayesian classifier to classify text against a set of trained categories it does so by generating a score (as a Float) for each possible category. The higher the score the closer the fit your text has with that category. The category with the highest score is returned as the best matching category.
132
-
133
- In *ClassifierReborn* the methods *classifications* and *classify_with_score* give you access to the calculated scores. The method *classify* only returns the best matching category.
134
-
135
- Knowing the score allows you to do some interesting things. For example if your application is to generate tags for a blog post you could use the *classifications* method to get a hash of the categories and their scores. You would sort on score and take only the top 3 or 4 categories as your tags for the blog post.
136
-
137
- You could within your application establish the smallest acceptable score and only use those categories whose score is greater than or equal to your smallest acceptable score as your tags for the blog post.
138
-
139
- But what if you only use the *classify* method? It does not show you the score of the best category. How do you know that the best category is really any good?
45
+ There is much more that can be done using Bayes and LSI beyond these quick examples.
46
+ For more information read the following documentation topics.
140
47
 
141
- You can use the threshold.
48
+ * [Installation and Dependencies](http://www.classifier-reborn.com/)
49
+ * [Bayesian Classifier](http://www.classifier-reborn.com/bayes)
50
+ * [Latent Semantic Indexer (LSI)](http://www.classifier-reborn.com/lsi)
51
+ * [Classifier Validation](http://www.classifier-reborn.com/validation)
52
+ * [Development and Contributions](http://www.classifier-reborn.com/development) (*Optional Docker instructions included*)
142
53
 
143
- #### Using the Threshold
144
-
145
- Some applications can have only one category. The application wants to know if the text being classified is of that category or not. For example consider a list of normal free text responses to some question or maybe a URL string coming to your web application. You know what a normal response looks like; but, you have no idea how people might mis-use the response. So what you want to do is create a bayesian classifier that just has one category, for example 'Good' and you want to know wither your text is classified as Good or Not Good.
146
-
147
- Or suppose you just want the ability to have multiple categories and a 'None of the Above' as a possibility.
148
-
149
- ##### Threshold
150
-
151
- When you initialize the *ClassifierReborn::Bayes* classifier there are several options which can be set that control threshold processing.
152
-
153
- ```ruby
154
- b = ClassifierReborn::Bayes.new(
155
- 'good', # one or more categories
156
- enable_threshold: true, # default: false
157
- threshold: -10.0 # default: 0.0
158
- )
159
- b.train_good 'good stuff from Dobie Gillis'
160
- # ...
161
- text = 'bad junk from Maynard G. Krebs'
162
- result = b.classify text
163
- if result.nil?
164
- STDERR.puts "ALERT: This is not good: #{text}"
165
- let_loose_the_dogs_of_war! # method definition left to the reader
166
- end
167
-
168
- ```
169
-
170
- In the *classify* method when the best category for the text has a score that is either less than the established threshold or is Float::INIFINITY, a nil category is returned. When you see a nil value returned from the *classify* method it means that none of the trained categories (regardless or how many categories were trained) has a score that is above or equal to the established threshold.
171
-
172
- #### Other Threshold-related Convience Methods
173
-
174
- ```ruby
175
- b.threshold # get the current threshold
176
- b.threshold = -10.0 # set the threshold
177
- b.threshold_enabled? # Boolean: is the threshold enabled?
178
- b.threshold_disabled? # Boolean: is the threshold disabled?
179
- b.enable_threshold # enables threshold processing
180
- b.disable_threshold # disables threshold processing
181
- ```
182
-
183
- Using these convience methods your applications can dynamically adjust threshold processing as required.
184
-
185
- ### Bayesian Classification
186
-
187
- * https://en.wikipedia.org/wiki/Naive_Bayes_classifier
188
- * http://www.process.com/precisemail/bayesian_filtering.htm
189
- * http://en.wikipedia.org/wiki/Bayesian_filtering
190
- * http://www.paulgraham.com/spam.html
191
-
192
- ## LSI
193
-
194
- A Latent Semantic Indexer by David Fayram. Latent Semantic Indexing engines
195
- are not as fast or as small as Bayesian classifiers, but are more flexible, providing
196
- fast search and clustering detection as well as semantic analysis of the text that
197
- theoretically simulates human learning.
198
-
199
- ### Usage
54
+ ### Notes on JRuby support
200
55
 
201
56
  ```ruby
202
- require 'classifier-reborn'
203
- lsi = ClassifierReborn::LSI.new
204
- strings = [ ["This text deals with dogs. Dogs.", :dog],
205
- ["This text involves dogs too. Dogs! ", :dog],
206
- ["This text revolves around cats. Cats.", :cat],
207
- ["This text also involves cats. Cats!", :cat],
208
- ["This text involves birds. Birds.",:bird ]]
209
- strings.each {|x| lsi.add_item x.first, x.last}
210
-
211
- lsi.search("dog", 3)
212
- # returns => ["This text deals with dogs. Dogs.", "This text involves dogs too. Dogs! ",
213
- # "This text also involves cats. Cats!"]
214
-
215
- lsi.find_related(strings[2], 2)
216
- # returns => ["This text revolves around cats. Cats.", "This text also involves cats. Cats!"]
217
-
218
- lsi.classify "This text is also about dogs!"
219
- # returns => :dog
57
+ gem 'classifier-reborn-jruby', platforms: :java
220
58
  ```
221
59
 
222
- Please see the ClassifierReborn::LSI documentation for more information. It is possible to index, search and classify
223
- with more than just simple strings.
224
-
225
- ### Latent Semantic Indexing
226
-
227
- * http://www.c2.com/cgi/wiki?LatentSemanticIndexing
228
- * http://www.chadfowler.com/index.cgi/Computing/LatentSemanticIndexing.rdoc
229
- * http://en.wikipedia.org/wiki/Latent_semantic_analysis
60
+ While experimental, this gem should work on JRuby without any kind of additional changes. Unfortunately, you will **not** be able to use C bindings to GNU/GSL or similar performance-enhancing native code. Additionally, we do not use `fast_stemmer`, but rather [an implementation](https://tartarus.org/martin/PorterStemmer/java.txt) of the [Porter Stemming](https://tartarus.org/martin/PorterStemmer/) algorithm. Stemming will differ between MRI and JRuby, however you may choose to [disable stemming](https://tartarus.org/martin/PorterStemmer/) and do your own manual preprocessing (or use some other [popular Java library](https://opennlp.apache.org/)).
230
61
 
62
+ If you encounter a problem, please submit your issue with `[JRuby]` in the title.
231
63
 
232
64
  ## Code of Conduct
233
65
 
234
- In order to have a more open and welcoming community, Classifier-Reborn adheres to the Jekyll
235
- [code of conduct](https://github.com/jekyll/jekyll/blob/master/CONDUCT.markdown) adapted from the Ruby on Rails code of
236
- conduct.
237
-
238
- Please adhere to this code of conduct in any interactions you have in the
239
- Classifier community. If you encounter someone violating
240
- these terms, please let [@chase](https://github.com/Ch4s3) know and we will address it as soon as possible.
66
+ In order to have a more open and welcoming community, `Classifier Reborn` adheres to the `Jekyll`
67
+ [code of conduct](https://github.com/jekyll/jekyll/blob/master/CODE_OF_CONDUCT.markdown) adapted from the `Ruby on Rails` code of conduct.
241
68
 
69
+ Please adhere to this code of conduct in any interactions you have in the `Classifier` community.
70
+ If you encounter someone violating these terms, please let [Chase Gilliam](https://github.com/Ch4s3) know and we will address it as soon as possible.
242
71
 
243
- ## Authors
72
+ ## Authors and Contributors
244
73
 
245
- * Lucas Carlson (lucas@rufy.com)
246
- * David Fayram II (dfayram@gmail.com)
247
- * Cameron McBride (cameron.mcbride@gmail.com)
248
- * Ivan Acosta-Rubio (ivan@softwarecriollo.com)
249
- * Parker Moore (email@byparker.com)
250
- * Chase Gilliam (chase.gilliam@gmail.com)
74
+ * [Lucas Carlson](mailto:lucas@rufy.com)
75
+ * [David Fayram II](mailto:dfayram@gmail.com)
76
+ * [Cameron McBride](mailto:cameron.mcbride@gmail.com)
77
+ * [Ivan Acosta-Rubio](mailto:ivan@softwarecriollo.com)
78
+ * [Parker Moore](mailto:email@byparker.com)
79
+ * [Chase Gilliam](mailto:chase.gilliam@gmail.com)
80
+ * and [many more](https://github.com/jekyll/classifier-reborn/graphs/contributors)...
251
81
 
252
- This library is released under the terms of the GNU LGPL. See LICENSE for more details.
82
+ The Classifier Reborn library is released under the terms of the [GNU LGPL-2.1](https://github.com/jekyll/classifier-reborn/blob/master/LICENSE).
@@ -0,0 +1,104 @@
1
+
2
+ فى
3
+ في
4
+ كل
5
+ لم
6
+ لن
7
+ له
8
+ من
9
+ هو
10
+ هي
11
+ قوة
12
+ كما
13
+ لها
14
+ منذ
15
+ وقد
16
+ ولا
17
+ لقاء
18
+ مقابل
19
+ هناك
20
+ وقال
21
+ وكان
22
+ وقالت
23
+ وكانت
24
+ فيه
25
+ لكن
26
+ وفي
27
+ ولم
28
+ ومن
29
+ وهو
30
+ وهي
31
+ يوم
32
+ فيها
33
+ منها
34
+ يكون
35
+ يمكن حيث
36
+ االا
37
+ اما
38
+ االتى
39
+ التي
40
+ اكثر
41
+ ايضا
42
+ الذى
43
+ الذي
44
+ الان
45
+ الذين
46
+ ابين
47
+ ذلك
48
+ دون
49
+ حول
50
+ حين
51
+ الى
52
+ انه
53
+ اول
54
+ انها
55
+ ف
56
+ و
57
+ و6
58
+ قد
59
+ لا
60
+ ما
61
+ مع
62
+ هذا
63
+ واحد
64
+ واضاف
65
+ واضافت
66
+ فان
67
+ قبل
68
+ قال
69
+ كان
70
+ لدى
71
+ نحو
72
+ هذه
73
+ وان
74
+ واكد
75
+ كانت
76
+ واوضح
77
+ ب
78
+ ا
79
+ أ
80
+ ،
81
+ عن
82
+ عند
83
+ عندما
84
+ على
85
+ عليه
86
+ عليها
87
+ تم
88
+ ضد
89
+ بعد
90
+ بعض
91
+ حتى
92
+ اذا
93
+ احد
94
+ بان
95
+ اجل
96
+ غير
97
+ بن
98
+ به
99
+ ثم
100
+ اف
101
+ ان
102
+ او
103
+ اي
104
+ بها