wikipedia-vandalism_detection 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +17 -0
- data/Gemfile +6 -0
- data/LICENSE.txt +4 -0
- data/README.md +265 -0
- data/Rakefile +12 -0
- data/lib/java/LibSVM.jar +0 -0
- data/lib/java/SMOTE.jar +0 -0
- data/lib/java/balancedRandomForest.jar +0 -0
- data/lib/java/diffutils-1.3.0.jar +0 -0
- data/lib/java/libsvm.jar +0 -0
- data/lib/java/oneClassClassifier.jar +0 -0
- data/lib/java/realAdaBoost.jar +0 -0
- data/lib/java/swc-engine-1.1.0-jar-with-dependencies.jar +0 -0
- data/lib/java/sweble-wikitext-extractor.jar +0 -0
- data/lib/weka/classifiers/functions/lib_svm.rb +15 -0
- data/lib/weka/classifiers/meta/one_class_classifier.rb +25 -0
- data/lib/weka/classifiers/meta/real_ada_boost.rb +17 -0
- data/lib/weka/classifiers/trees/balanced_random_forest.rb +18 -0
- data/lib/weka/filters/supervised/instance/smote.rb +22 -0
- data/lib/wikipedia.rb +51 -0
- data/lib/wikipedia/vandalism_detection.rb +30 -0
- data/lib/wikipedia/vandalism_detection/algorithms.rb +18 -0
- data/lib/wikipedia/vandalism_detection/algorithms/kullback_leibler_divergence.rb +69 -0
- data/lib/wikipedia/vandalism_detection/classifier.rb +186 -0
- data/lib/wikipedia/vandalism_detection/configuration.rb +321 -0
- data/lib/wikipedia/vandalism_detection/diff.rb +27 -0
- data/lib/wikipedia/vandalism_detection/edit.rb +75 -0
- data/lib/wikipedia/vandalism_detection/evaluator.rb +606 -0
- data/lib/wikipedia/vandalism_detection/exceptions.rb +40 -0
- data/lib/wikipedia/vandalism_detection/feature_calculator.rb +89 -0
- data/lib/wikipedia/vandalism_detection/features.rb +67 -0
- data/lib/wikipedia/vandalism_detection/features/all_wordlists_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/all_wordlists_impact.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/anonymity.rb +19 -0
- data/lib/wikipedia/vandalism_detection/features/anonymity_previous.rb +29 -0
- data/lib/wikipedia/vandalism_detection/features/article_size.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/bad_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/bad_impact.rb +21 -0
- data/lib/wikipedia/vandalism_detection/features/base.rb +54 -0
- data/lib/wikipedia/vandalism_detection/features/biased_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/biased_impact.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/blanking.rb +25 -0
- data/lib/wikipedia/vandalism_detection/features/character_diversity.rb +25 -0
- data/lib/wikipedia/vandalism_detection/features/character_sequence.rb +19 -0
- data/lib/wikipedia/vandalism_detection/features/comment_bad_frequency.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/comment_biased_frequency.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/comment_length.rb +17 -0
- data/lib/wikipedia/vandalism_detection/features/comment_markup_frequency.rb +27 -0
- data/lib/wikipedia/vandalism_detection/features/comment_pronoun_frequency.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/comment_sex_frequency.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/comment_vulgarism_frequency.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/compressibility.rb +27 -0
- data/lib/wikipedia/vandalism_detection/features/contains_base.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/copyedit.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/digit_ratio.rb +24 -0
- data/lib/wikipedia/vandalism_detection/features/edits_per_user.rb +65 -0
- data/lib/wikipedia/vandalism_detection/features/emoticons_frequency.rb +27 -0
- data/lib/wikipedia/vandalism_detection/features/emoticons_impact.rb +29 -0
- data/lib/wikipedia/vandalism_detection/features/frequency_base.rb +20 -0
- data/lib/wikipedia/vandalism_detection/features/impact_base.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/inserted_character_distribution.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/inserted_external_links.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/inserted_internal_links.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/inserted_size.rb +20 -0
- data/lib/wikipedia/vandalism_detection/features/inserted_words.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/longest_word.rb +20 -0
- data/lib/wikipedia/vandalism_detection/features/markup_frequency.rb +27 -0
- data/lib/wikipedia/vandalism_detection/features/markup_impact.rb +27 -0
- data/lib/wikipedia/vandalism_detection/features/non_alphanumeric_ratio.rb +24 -0
- data/lib/wikipedia/vandalism_detection/features/personal_life.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/pronoun_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/pronoun_impact.rb +21 -0
- data/lib/wikipedia/vandalism_detection/features/removed_all_wordlists_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/removed_bad_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/removed_biased_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/removed_character_distribution.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/removed_emoticons_frequency.rb +27 -0
- data/lib/wikipedia/vandalism_detection/features/removed_markup_frequency.rb +28 -0
- data/lib/wikipedia/vandalism_detection/features/removed_pronoun_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/removed_sex_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/removed_size.rb +20 -0
- data/lib/wikipedia/vandalism_detection/features/removed_vulgarism_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/removed_words.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/replacement_similarity.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/reverted.rb +18 -0
- data/lib/wikipedia/vandalism_detection/features/revisions_character_distribution.rb +21 -0
- data/lib/wikipedia/vandalism_detection/features/same_editor.rb +29 -0
- data/lib/wikipedia/vandalism_detection/features/sex_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/sex_impact.rb +21 -0
- data/lib/wikipedia/vandalism_detection/features/size_increment.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/size_ratio.rb +26 -0
- data/lib/wikipedia/vandalism_detection/features/term_frequency.rb +25 -0
- data/lib/wikipedia/vandalism_detection/features/time_interval.rb +31 -0
- data/lib/wikipedia/vandalism_detection/features/time_of_day.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/upper_case_ratio.rb +24 -0
- data/lib/wikipedia/vandalism_detection/features/upper_case_words_ratio.rb +31 -0
- data/lib/wikipedia/vandalism_detection/features/upper_to_lower_case_ratio.rb +24 -0
- data/lib/wikipedia/vandalism_detection/features/user_reputation.rb +38 -0
- data/lib/wikipedia/vandalism_detection/features/vulgarism_frequency.rb +23 -0
- data/lib/wikipedia/vandalism_detection/features/vulgarism_impact.rb +22 -0
- data/lib/wikipedia/vandalism_detection/features/weekday.rb +21 -0
- data/lib/wikipedia/vandalism_detection/features/words_increment.rb +22 -0
- data/lib/wikipedia/vandalism_detection/instances.rb +130 -0
- data/lib/wikipedia/vandalism_detection/page.rb +88 -0
- data/lib/wikipedia/vandalism_detection/page_parser.rb +52 -0
- data/lib/wikipedia/vandalism_detection/revision.rb +69 -0
- data/lib/wikipedia/vandalism_detection/revision_parser.rb +43 -0
- data/lib/wikipedia/vandalism_detection/test_dataset.rb +367 -0
- data/lib/wikipedia/vandalism_detection/text.rb +18 -0
- data/lib/wikipedia/vandalism_detection/training_dataset.rb +303 -0
- data/lib/wikipedia/vandalism_detection/version.rb +5 -0
- data/lib/wikipedia/vandalism_detection/wikitext_extractor.rb +80 -0
- data/lib/wikipedia/vandalism_detection/word_lists.rb +19 -0
- data/lib/wikipedia/vandalism_detection/word_lists/bad.rb +12 -0
- data/lib/wikipedia/vandalism_detection/word_lists/biased.rb +21 -0
- data/lib/wikipedia/vandalism_detection/word_lists/emoticons.rb +22 -0
- data/lib/wikipedia/vandalism_detection/word_lists/markup.rb +12 -0
- data/lib/wikipedia/vandalism_detection/word_lists/pronouns.rb +15 -0
- data/lib/wikipedia/vandalism_detection/word_lists/sex.rb +12 -0
- data/lib/wikipedia/vandalism_detection/word_lists/vulgarism.rb +97 -0
- data/spec/factories/edit.rb +20 -0
- data/spec/factories/page.rb +13 -0
- data/spec/factories/revision.rb +51 -0
- data/spec/resources/config/config.yml +35 -0
- data/spec/resources/corpora/test/edits.csv +8 -0
- data/spec/resources/corpora/test/ground-truth.txt +3 -0
- data/spec/resources/corpora/test/revisions/part-1/326471754.txt +199 -0
- data/spec/resources/corpora/test/revisions/part-1/326873205.txt +137 -0
- data/spec/resources/corpora/test/revisions/part-1/328774035.txt +162 -0
- data/spec/resources/corpora/test/revisions/part-2/307084144.txt +137 -0
- data/spec/resources/corpora/test/revisions/part-2/326978767.txt +199 -0
- data/spec/resources/corpora/test/revisions/part-2/328774110.txt +162 -0
- data/spec/resources/corpora/training/annotations.csv +7 -0
- data/spec/resources/corpora/training/edits.csv +7 -0
- data/spec/resources/corpora/training/revisions/part-1/326471754.txt +199 -0
- data/spec/resources/corpora/training/revisions/part-1/326873205.txt +137 -0
- data/spec/resources/corpora/training/revisions/part-1/328774035.txt +162 -0
- data/spec/resources/corpora/training/revisions/part-2/307084144.txt +137 -0
- data/spec/resources/corpora/training/revisions/part-2/326978767.txt +199 -0
- data/spec/resources/corpora/training/revisions/part-2/328774110.txt +162 -0
- data/spec/resources/page_with_redirects.xml +85 -0
- data/spec/resources/redirect_page.xml +59 -0
- data/spec/resources/revision_simplified.xml +13 -0
- data/spec/resources/sample_revision.txt +137 -0
- data/spec/resources/sample_revision_clean_text.txt +1 -0
- data/spec/resources/sample_revision_plain_text.txt +183 -0
- data/spec/resources/vandalism_on_wikipedia.xml +234 -0
- data/spec/resources/vandalism_on_wikipedia_simplified.xml +119 -0
- data/spec/resources/wikipedia_tokens.txt +30 -0
- data/spec/spec_helper.rb +22 -0
- data/spec/support/macros/file_reading.rb +7 -0
- data/spec/support/macros/test_configuration.rb +71 -0
- data/spec/vandalism_detection/algorithms/kullback_leibler_divergence_spec.rb +36 -0
- data/spec/vandalism_detection/classifier_spec.rb +317 -0
- data/spec/vandalism_detection/configuration_spec.rb +517 -0
- data/spec/vandalism_detection/diff_spec.rb +40 -0
- data/spec/vandalism_detection/edit_spec.rb +137 -0
- data/spec/vandalism_detection/evaluator_spec.rb +671 -0
- data/spec/vandalism_detection/feature_calculator_spec.rb +128 -0
- data/spec/vandalism_detection/features/all_wordlists_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/all_wordlists_impact_spec.rb +58 -0
- data/spec/vandalism_detection/features/anonymity_previous_spec.rb +61 -0
- data/spec/vandalism_detection/features/anonymity_spec.rb +23 -0
- data/spec/vandalism_detection/features/article_size_spec.rb +35 -0
- data/spec/vandalism_detection/features/bad_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/bad_impact_spec.rb +59 -0
- data/spec/vandalism_detection/features/base_spec.rb +49 -0
- data/spec/vandalism_detection/features/biased_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/biased_impact_spec.rb +58 -0
- data/spec/vandalism_detection/features/blanking_spec.rb +38 -0
- data/spec/vandalism_detection/features/character_diversity_spec.rb +35 -0
- data/spec/vandalism_detection/features/character_sequence_spec.rb +37 -0
- data/spec/vandalism_detection/features/comment_bad_frequency_spec.rb +34 -0
- data/spec/vandalism_detection/features/comment_biased_frequency_spec.rb +34 -0
- data/spec/vandalism_detection/features/comment_length_spec.rb +27 -0
- data/spec/vandalism_detection/features/comment_markup_frequency_spec.rb +34 -0
- data/spec/vandalism_detection/features/comment_pronoun_frequency_spec.rb +34 -0
- data/spec/vandalism_detection/features/comment_sex_frequency_spec.rb +34 -0
- data/spec/vandalism_detection/features/comment_vulgarism_frequency_spec.rb +34 -0
- data/spec/vandalism_detection/features/compressibility_spec.rb +42 -0
- data/spec/vandalism_detection/features/contains_base_spec.rb +33 -0
- data/spec/vandalism_detection/features/copyedit_spec.rb +33 -0
- data/spec/vandalism_detection/features/digit_ratio_spec.rb +35 -0
- data/spec/vandalism_detection/features/edits_per_user_spec.rb +49 -0
- data/spec/vandalism_detection/features/emoticons_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/emoticons_impact_spec.rb +51 -0
- data/spec/vandalism_detection/features/frequency_base_spec.rb +26 -0
- data/spec/vandalism_detection/features/impact_base_spec.rb +41 -0
- data/spec/vandalism_detection/features/inserted_character_distribution_spec.rb +46 -0
- data/spec/vandalism_detection/features/inserted_external_links_spec.rb +35 -0
- data/spec/vandalism_detection/features/inserted_internal_links_spec.rb +35 -0
- data/spec/vandalism_detection/features/inserted_size_spec.rb +35 -0
- data/spec/vandalism_detection/features/inserted_words_spec.rb +35 -0
- data/spec/vandalism_detection/features/longest_word_spec.rb +35 -0
- data/spec/vandalism_detection/features/markup_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/markup_impact_spec.rb +59 -0
- data/spec/vandalism_detection/features/non_alphanumeric_ratio_spec.rb +35 -0
- data/spec/vandalism_detection/features/personal_life_spec.rb +26 -0
- data/spec/vandalism_detection/features/pronoun_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/pronoun_impact_spec.rb +59 -0
- data/spec/vandalism_detection/features/removed_all_wordlists_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/removed_bad_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/removed_biased_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/removed_character_distribution_spec.rb +46 -0
- data/spec/vandalism_detection/features/removed_emoticons_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/removed_markup_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/removed_pronoun_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/removed_sex_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/removed_size_spec.rb +35 -0
- data/spec/vandalism_detection/features/removed_vulgarism_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/removed_words_spec.rb +35 -0
- data/spec/vandalism_detection/features/replacement_similarity_spec.rb +44 -0
- data/spec/vandalism_detection/features/reverted_spec.rb +28 -0
- data/spec/vandalism_detection/features/revisions_character_distribution_spec.rb +46 -0
- data/spec/vandalism_detection/features/same_editor_spec.rb +60 -0
- data/spec/vandalism_detection/features/sex_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/sex_impact_spec.rb +59 -0
- data/spec/vandalism_detection/features/size_increment_spec.rb +35 -0
- data/spec/vandalism_detection/features/size_ratio_spec.rb +57 -0
- data/spec/vandalism_detection/features/term_frequency_spec.rb +38 -0
- data/spec/vandalism_detection/features/time_interval_spec.rb +50 -0
- data/spec/vandalism_detection/features/time_of_day_spec.rb +22 -0
- data/spec/vandalism_detection/features/upper_case_ratio_spec.rb +35 -0
- data/spec/vandalism_detection/features/upper_case_words_ratio_spec.rb +37 -0
- data/spec/vandalism_detection/features/upper_to_lower_case_ratio_spec.rb +35 -0
- data/spec/vandalism_detection/features/user_reputation_spec.rb +52 -0
- data/spec/vandalism_detection/features/vulgarism_frequency_spec.rb +36 -0
- data/spec/vandalism_detection/features/vulgarism_impact_spec.rb +58 -0
- data/spec/vandalism_detection/features/weekday_spec.rb +22 -0
- data/spec/vandalism_detection/features/words_increment_spec.rb +35 -0
- data/spec/vandalism_detection/instances_spec.rb +156 -0
- data/spec/vandalism_detection/page_parser_spec.rb +184 -0
- data/spec/vandalism_detection/page_spec.rb +135 -0
- data/spec/vandalism_detection/revision_parser_spec.rb +53 -0
- data/spec/vandalism_detection/revision_spec.rb +115 -0
- data/spec/vandalism_detection/test_dataset_spec.rb +231 -0
- data/spec/vandalism_detection/text_spec.rb +29 -0
- data/spec/vandalism_detection/training_dataset_spec.rb +264 -0
- data/spec/vandalism_detection/wikitext_extractor_spec.rb +72 -0
- data/spec/weka/classifiers/functions/lib_svm_spec.rb +38 -0
- data/spec/weka/classifiers/meta/one_class_classifier_spec.rb +76 -0
- data/spec/weka/classifiers/meta/real_ada_boost_spec.rb +40 -0
- data/spec/weka/classifiers/trees/balanced_random_forest_spec.rb +40 -0
- data/spec/weka/filters/supervised/instance/smote_spec.rb +6 -0
- data/wikipedia-vandalism_detection.gemspec +30 -0
- metadata +512 -0
@@ -0,0 +1,36 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::SexFrequency do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::SexFrequency.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::FrequencyBase }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns the number of vulgarism words relative to all words count" do
|
14
|
+
# total 6 words, 3 bad
|
15
|
+
old_text = Wikipedia::VandalismDetection::Text.new 'Old whatever.'
|
16
|
+
new_text = Wikipedia::VandalismDetection::Text.new "Old whatever. New sex contents. Penis, dildos, boy."
|
17
|
+
|
18
|
+
old_revision = build(:old_revision, text: old_text)
|
19
|
+
new_revision = build(:new_revision, text: new_text)
|
20
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
21
|
+
|
22
|
+
expect(@feature.calculate(edit)).to eq 3.0 / 6.0
|
23
|
+
end
|
24
|
+
|
25
|
+
it "returns 0.0 on emtpy clean text revisions" do
|
26
|
+
old_text = Wikipedia::VandalismDetection::Text.new 'Old guy.'
|
27
|
+
new_text = Wikipedia::VandalismDetection::Text.new "Old guy. {{speedy deletion}}"
|
28
|
+
|
29
|
+
old_revision = build(:old_revision, text: old_text)
|
30
|
+
new_revision = build(:new_revision, text: new_text)
|
31
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
32
|
+
|
33
|
+
expect(@feature.calculate(edit)).to eq 0.0
|
34
|
+
end
|
35
|
+
end
|
36
|
+
end
|
@@ -0,0 +1,59 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::SexImpact do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::SexImpact.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::ImpactBase }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns the impact of sex words of the edit's new revision text" do
|
14
|
+
# 3 sex words
|
15
|
+
old_text = Wikipedia::VandalismDetection::Text.new "Penis, old text dildo, breast it"
|
16
|
+
|
17
|
+
# 4 sex words
|
18
|
+
new_text = Wikipedia::VandalismDetection::Text.new "Penis, old text dildo, breast anal it"
|
19
|
+
|
20
|
+
old_revision = build(:old_revision, text: old_text)
|
21
|
+
new_revision = build(:new_revision, text: new_text)
|
22
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
23
|
+
|
24
|
+
expect(@feature.calculate(edit)).to eq 3.0 / (3.0 + 4.0)
|
25
|
+
end
|
26
|
+
|
27
|
+
it "returns 0.5 on both no terms in text revisions" do
|
28
|
+
text = Wikipedia::VandalismDetection::Text.new "{{speedy deletion}}"
|
29
|
+
|
30
|
+
old_revision = build(:old_revision, text: text)
|
31
|
+
new_revision = build(:new_revision, text: text)
|
32
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
33
|
+
|
34
|
+
expect(@feature.calculate(edit)).to eq 0.5
|
35
|
+
end
|
36
|
+
|
37
|
+
it "returns 0.0 on emtpy clean text of old revision" do
|
38
|
+
old_text = Wikipedia::VandalismDetection::Text.new "{{speedy deletion}}"
|
39
|
+
new_text = Wikipedia::VandalismDetection::Text.new "anal"
|
40
|
+
|
41
|
+
old_revision = build(:old_revision, text: old_text)
|
42
|
+
new_revision = build(:new_revision, text: new_text)
|
43
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
44
|
+
|
45
|
+
expect(@feature.calculate(edit)).to eq 0.0
|
46
|
+
end
|
47
|
+
|
48
|
+
it "returns 1.0 on emtpy clean text of new revision" do
|
49
|
+
old_text = Wikipedia::VandalismDetection::Text.new "anal"
|
50
|
+
new_text = Wikipedia::VandalismDetection::Text.new "{{speedy deletion}}"
|
51
|
+
|
52
|
+
old_revision = build(:old_revision, text: old_text)
|
53
|
+
new_revision = build(:new_revision, text: new_text)
|
54
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
55
|
+
|
56
|
+
expect(@feature.calculate(edit)).to eq 1.0
|
57
|
+
end
|
58
|
+
end
|
59
|
+
end
|
@@ -0,0 +1,35 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::SizeIncrement do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::SizeIncrement.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::Base }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns a negative increment on more removed texts" do
|
14
|
+
old_revision_text = Wikipedia::VandalismDetection::Text.new '123456789' # length 9
|
15
|
+
new_revision_text = Wikipedia::VandalismDetection::Text.new '123' # length 3
|
16
|
+
|
17
|
+
old_revision = build(:old_revision, text: old_revision_text)
|
18
|
+
new_revision = build(:new_revision, text: new_revision_text)
|
19
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
20
|
+
|
21
|
+
expect(@feature.calculate(edit)).to eq 3.0 - 9.0
|
22
|
+
end
|
23
|
+
|
24
|
+
it "returns a positive increment on more removed texts" do
|
25
|
+
old_revision_text = Wikipedia::VandalismDetection::Text.new '123' # length 3
|
26
|
+
new_revision_text = Wikipedia::VandalismDetection::Text.new '123456789' # length 9
|
27
|
+
|
28
|
+
old_revision = build(:old_revision, text: old_revision_text)
|
29
|
+
new_revision = build(:new_revision, text: new_revision_text)
|
30
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
31
|
+
|
32
|
+
expect(@feature.calculate(edit)).to eq 9.0 - 3.0
|
33
|
+
end
|
34
|
+
end
|
35
|
+
end
|
@@ -0,0 +1,57 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::SizeRatio do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::SizeRatio.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::Base }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns the ratio of the edit's revisions text sizes" do
|
14
|
+
old_revision_text = Wikipedia::VandalismDetection::Text.new '123456789' # length 9
|
15
|
+
new_revision_text = Wikipedia::VandalismDetection::Text.new '123' # length 3
|
16
|
+
|
17
|
+
old_revision = build(:old_revision, text: old_revision_text)
|
18
|
+
new_revision = build(:new_revision, text: new_revision_text)
|
19
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
20
|
+
|
21
|
+
expect(@feature.calculate(edit)).to eq 9.0 / (9.0 + 3.0)
|
22
|
+
end
|
23
|
+
|
24
|
+
it "returns 1.0 on emtpy new text revisions" do
|
25
|
+
old_text = Wikipedia::VandalismDetection::Text.new "sample text"
|
26
|
+
new_text = Wikipedia::VandalismDetection::Text.new ""
|
27
|
+
|
28
|
+
old_revision = build(:old_revision, text: old_text)
|
29
|
+
new_revision = build(:new_revision, text: new_text)
|
30
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
31
|
+
|
32
|
+
expect(@feature.calculate(edit)).to eq 1.0
|
33
|
+
end
|
34
|
+
|
35
|
+
it "returns 0.0 on emtpy old text revisions" do
|
36
|
+
old_text = Wikipedia::VandalismDetection::Text.new ""
|
37
|
+
new_text = Wikipedia::VandalismDetection::Text.new "sample text"
|
38
|
+
|
39
|
+
old_revision = build(:old_revision, text: old_text)
|
40
|
+
new_revision = build(:new_revision, text: new_text)
|
41
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
42
|
+
|
43
|
+
expect(@feature.calculate(edit)).to eq 0.0
|
44
|
+
end
|
45
|
+
|
46
|
+
it "returns 0.5 on emtpy both text revisions" do
|
47
|
+
old_text = Wikipedia::VandalismDetection::Text.new ""
|
48
|
+
new_text = Wikipedia::VandalismDetection::Text.new ""
|
49
|
+
|
50
|
+
old_revision = build(:old_revision, text: old_text)
|
51
|
+
new_revision = build(:new_revision, text: new_text)
|
52
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
53
|
+
|
54
|
+
expect(@feature.calculate(edit)).to eq 0.5
|
55
|
+
end
|
56
|
+
end
|
57
|
+
end
|
@@ -0,0 +1,38 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::TermFrequency do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::TermFrequency.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::FrequencyBase }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns the average relative frequency of inserted words in the new revision" do
|
14
|
+
# removed [you, I], added [you, you, you, we]
|
15
|
+
# for [you, we] compute frequency in old_text and average for all words
|
16
|
+
# here: [you] is 6x in new text, [we] is 3x in new text of total 10 words
|
17
|
+
# avrg = (6/10 + 3/10)/2
|
18
|
+
old_text = Wikipedia::VandalismDetection::Text.new "we\nwe\nyou\nyou\nyou\nI\nyou\nI\n"
|
19
|
+
new_text = Wikipedia::VandalismDetection::Text.new "we\nwe\nyou\nyou\nI\nyou\n''(you''\nyou\nyou\n[[we]])\n"
|
20
|
+
|
21
|
+
old_revision = build(:old_revision, text: old_text)
|
22
|
+
new_revision = build(:new_revision, text: new_text)
|
23
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
24
|
+
|
25
|
+
expect(@feature.calculate(edit)).to eq ((6.0/10.0) + (3.0/10.0)) / 2.0
|
26
|
+
end
|
27
|
+
|
28
|
+
it "returns 0.0 on emtpy clean text revisions" do
|
29
|
+
text = Wikipedia::VandalismDetection::Text.new "{{speedy deletion}}"
|
30
|
+
|
31
|
+
old_revision = build(:old_revision, text: text)
|
32
|
+
new_revision = build(:new_revision, text: text)
|
33
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
34
|
+
|
35
|
+
expect(@feature.calculate(edit)).to eq 0.0
|
36
|
+
end
|
37
|
+
end
|
38
|
+
end
|
@@ -0,0 +1,50 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::TimeInterval do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::TimeInterval.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::Base }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns time interval in days from old to new revision" do
|
14
|
+
old_revision_time = '2014-11-27T18:00:00Z'
|
15
|
+
new_revision_time = '2014-11-29T06:00:00Z'
|
16
|
+
|
17
|
+
old_revision = build(:old_revision, timestamp: old_revision_time)
|
18
|
+
new_revision = build(:new_revision, timestamp: new_revision_time)
|
19
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
20
|
+
|
21
|
+
expect(@feature.calculate(edit)).to eq 1.5
|
22
|
+
end
|
23
|
+
|
24
|
+
it "requests the time from Wikipedia API if old revisions timestamp is not given" do
|
25
|
+
# to get api call, see:
|
26
|
+
# https://en.wikipedia.org/w/api.php?action=query&prop=revisions&rvprop=timestamp&revids=327585467
|
27
|
+
# => 2009-11-24T01:57:35Z
|
28
|
+
new_revision_time = '2009-11-24T13:57:35Z'
|
29
|
+
|
30
|
+
old_revision = build(:old_revision, id: '327585467', timestamp: nil)
|
31
|
+
new_revision = build(:new_revision, id: '327607921', parent_id: '327585467', timestamp: new_revision_time)
|
32
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
33
|
+
|
34
|
+
expect(@feature.calculate(edit)).to eq 0.5
|
35
|
+
end
|
36
|
+
|
37
|
+
it "returns missing if old reivision is not available anymore" do
|
38
|
+
# to get api call, see:
|
39
|
+
# https://en.wikipedia.org/w/api.php?action=query&prop=revisions&rvprop=timestamp&revids=325218985
|
40
|
+
# <rev revid="325218985"/>
|
41
|
+
new_revision_time = '2011-11-11T01:00:00Z'
|
42
|
+
|
43
|
+
old_revision = build(:old_revision, id: '325218985', timestamp: nil)
|
44
|
+
new_revision = build(:new_revision, id: '326980599', parent_id: '325218985', timestamp: new_revision_time)
|
45
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
46
|
+
|
47
|
+
expect(@feature.calculate(edit)).to eq Wikipedia::VandalismDetection::Features::MISSING_VALUE
|
48
|
+
end
|
49
|
+
end
|
50
|
+
end
|
@@ -0,0 +1,22 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::TimeOfDay do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::TimeOfDay.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::Base }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns the time of day as decimal value (hours)" do
|
14
|
+
old_revision = build(:old_revision)
|
15
|
+
new_revision = build(:new_revision, timestamp: '2012-12-09T05:30:36Z' )
|
16
|
+
|
17
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
18
|
+
|
19
|
+
expect(@feature.calculate(edit)).to eq 5.6
|
20
|
+
end
|
21
|
+
end
|
22
|
+
end
|
@@ -0,0 +1,35 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::UpperCaseRatio do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::UpperCaseRatio.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::Base }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns the uppercase to all letters ratio of the edit's new revision cleaned inserted text" do
|
14
|
+
old_text = Wikipedia::VandalismDetection::Text.new('text')
|
15
|
+
new_text = Wikipedia::VandalismDetection::Text.new 'text [[1A 4B6 8Cd]]' # 3 uppercase letters of total 4 inserted letters
|
16
|
+
|
17
|
+
old_revision = build(:old_revision, text: old_text)
|
18
|
+
new_revision = build(:new_revision, text: new_text)
|
19
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
20
|
+
|
21
|
+
expect(@feature.calculate(edit)).to eq (1.0 + 3) / (1.0 + 4)
|
22
|
+
end
|
23
|
+
|
24
|
+
it "returns 0.0 if no text inserted" do
|
25
|
+
old_text = Wikipedia::VandalismDetection::Text.new("deletion text")
|
26
|
+
new_text = Wikipedia::VandalismDetection::Text.new("text")
|
27
|
+
|
28
|
+
old_revision = build(:old_revision, text: old_text)
|
29
|
+
new_revision = build(:new_revision, text: new_text)
|
30
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
31
|
+
|
32
|
+
expect(@feature.calculate(edit)).to eq 0.0
|
33
|
+
end
|
34
|
+
end
|
35
|
+
end
|
@@ -0,0 +1,37 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
require 'spec_helper'
|
3
|
+
|
4
|
+
describe Wikipedia::VandalismDetection::Features::UpperCaseWordsRatio do
|
5
|
+
|
6
|
+
before do
|
7
|
+
@feature = Wikipedia::VandalismDetection::Features::UpperCaseWordsRatio.new
|
8
|
+
end
|
9
|
+
|
10
|
+
it { should be_a Wikipedia::VandalismDetection::Features::Base }
|
11
|
+
|
12
|
+
describe "#calculate" do
|
13
|
+
|
14
|
+
it "returns the uppercase to all words ratio of the edit's new revision cleaned inserted text" do
|
15
|
+
old_text = Wikipedia::VandalismDetection::Text.new('text')
|
16
|
+
# 2 two capital (not numbers!) words of total 4 inserted, template {{23A}} is removed while cleaning!
|
17
|
+
new_text = Wikipedia::VandalismDetection::Text.new "text [[HELLO you]] NICE boyß3 1990 {{23A}}"
|
18
|
+
|
19
|
+
old_revision = build(:old_revision, text: old_text)
|
20
|
+
new_revision = build(:new_revision, text: new_text)
|
21
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
22
|
+
|
23
|
+
expect(@feature.calculate(edit)).to eq (1.0 + 2) / (1.0 + 4)
|
24
|
+
end
|
25
|
+
|
26
|
+
it "returns 0.0 if no text inserted" do
|
27
|
+
old_text = Wikipedia::VandalismDetection::Text.new("DELECTION text")
|
28
|
+
new_text = Wikipedia::VandalismDetection::Text.new("text")
|
29
|
+
|
30
|
+
old_revision = build(:old_revision, text: old_text)
|
31
|
+
new_revision = build(:new_revision, text: new_text)
|
32
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
33
|
+
|
34
|
+
expect(@feature.calculate(edit)).to eq 0.0
|
35
|
+
end
|
36
|
+
end
|
37
|
+
end
|
@@ -0,0 +1,35 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::UpperToLowerCaseRatio do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::UpperToLowerCaseRatio.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::Base }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns the uppercase to lowercase letters ratio of the edit's new revision inserted text" do
|
14
|
+
old_text = Wikipedia::VandalismDetection::Text.new('text')
|
15
|
+
new_text = Wikipedia::VandalismDetection::Text.new 'text [[1aA 4B6 8Cd ef]]' # 3 uppercase letters, 4 lowercase letters
|
16
|
+
|
17
|
+
old_revision = build(:old_revision, text: old_text)
|
18
|
+
new_revision = build(:new_revision, text: new_text)
|
19
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision)
|
20
|
+
|
21
|
+
expect(@feature.calculate(edit)).to eq (1.0 + 3) / (1.0 + 4)
|
22
|
+
end
|
23
|
+
|
24
|
+
it "returns 0.0 if no text inserted" do
|
25
|
+
old_text = Wikipedia::VandalismDetection::Text.new("deletion text")
|
26
|
+
new_text = Wikipedia::VandalismDetection::Text.new("text")
|
27
|
+
|
28
|
+
old_revision = build(:old_revision, text: old_text)
|
29
|
+
new_revision = build(:new_revision, text: new_text)
|
30
|
+
edit = build(:edit, new_revision: new_revision, old_revision: old_revision)
|
31
|
+
|
32
|
+
expect(@feature.calculate(edit)).to eq 0.0
|
33
|
+
end
|
34
|
+
end
|
35
|
+
end
|
@@ -0,0 +1,52 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
describe Wikipedia::VandalismDetection::Features::UserReputation do
|
4
|
+
|
5
|
+
before do
|
6
|
+
@feature = Wikipedia::VandalismDetection::Features::UserReputation.new
|
7
|
+
end
|
8
|
+
|
9
|
+
it { should be_a Wikipedia::VandalismDetection::Features::Base }
|
10
|
+
|
11
|
+
describe "#calculate" do
|
12
|
+
|
13
|
+
it "returns the average number of contributor's reputation (WikiTrust) in the new revision" do
|
14
|
+
#http://en.collaborativetrust.com/WikiTrust/RemoteAPI?method=wikimarkup&pageid=<page_id>&revid=<old_rev_id>
|
15
|
+
# here: http://en.collaborativetrust.com/WikiTrust/RemoteAPI?method=wikimarkup&pageid=15580374&revid=604074768
|
16
|
+
old_revision = build(:old_revision, id: '603070163')
|
17
|
+
new_revision = build(:new_revision, id: '604074768', parent_id: '603070163', contributor: 'David Levy')
|
18
|
+
page = Wikipedia::VandalismDetection::Page.new
|
19
|
+
page.id = "15580374"
|
20
|
+
|
21
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision, page: page)
|
22
|
+
|
23
|
+
expect(@feature.calculate(edit)).to eq 10.0
|
24
|
+
end
|
25
|
+
|
26
|
+
it "fetches the page id from the wikipedia API by page title if not available in edit" do
|
27
|
+
#http://en.collaborativetrust.com/WikiTrust/RemoteAPI?method=wikimarkup&pageid=<page_id>&revid=<old_rev_id>
|
28
|
+
# here: http://en.collaborativetrust.com/WikiTrust/RemoteAPI?method=wikimarkup&pageid=15580374&revid=604074768
|
29
|
+
old_revision = build(:old_revision, id: '603070163')
|
30
|
+
new_revision = build(:new_revision, id: '604074768', parent_id: '603070163', contributor: 'David Levy')
|
31
|
+
page = Wikipedia::VandalismDetection::Page.new
|
32
|
+
page.title = 'Main Page'
|
33
|
+
|
34
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision, page: page)
|
35
|
+
|
36
|
+
expect(@feature.calculate(edit)).to eq 10.0
|
37
|
+
end
|
38
|
+
|
39
|
+
it "returns 0.0 WikiTrust if no contributions are available for given user" do
|
40
|
+
#http://en.collaborativetrust.com/WikiTrust/RemoteAPI?method=wikimarkup&pageid=<page_id>&revid=<old_rev_id>
|
41
|
+
# here: http://en.collaborativetrust.com/WikiTrust/RemoteAPI?method=wikimarkup&pageid=15580374&revid=604074768
|
42
|
+
old_revision = build(:old_revision, id: '603070163')
|
43
|
+
new_revision = build(:new_revision, id: '604074768', parent_id: '603070163', contributor: 'Artificial User')
|
44
|
+
page = Wikipedia::VandalismDetection::Page.new
|
45
|
+
page.id = "15580374"
|
46
|
+
|
47
|
+
edit = build(:edit, old_revision: old_revision, new_revision: new_revision, page: page)
|
48
|
+
|
49
|
+
expect(@feature.calculate(edit)).to eq 0.0
|
50
|
+
end
|
51
|
+
end
|
52
|
+
end
|