svm_helper 0.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +7 -0
- data/.gitignore +22 -0
- data/.rspec +3 -0
- data/.ruby-version +1 -0
- data/.travis.yml +9 -0
- data/.versions.conf +4 -0
- data/.yardopts +3 -0
- data/Gemfile +24 -0
- data/Guardfile +17 -0
- data/LICENSE.txt +22 -0
- data/README.md +41 -0
- data/Rakefile +7 -0
- data/lib/svm_helper.rb +8 -0
- data/lib/svm_helper/feature_vector.rb +17 -0
- data/lib/svm_helper/interface_helper.rb +57 -0
- data/lib/svm_helper/preprocessed_data.rb +17 -0
- data/lib/svm_helper/preprocessors.rb +2 -0
- data/lib/svm_helper/preprocessors/simple.rb +111 -0
- data/lib/svm_helper/preprocessors/with_industry_map.rb +40 -0
- data/lib/svm_helper/selectors.rb +3 -0
- data/lib/svm_helper/selectors/n_gram.rb +31 -0
- data/lib/svm_helper/selectors/simple.rb +163 -0
- data/lib/svm_helper/selectors/with_binary_encoding.rb +42 -0
- data/lib/svm_helper/stopwords/de +127 -0
- data/lib/svm_helper/stopwords/en +119 -0
- data/lib/svm_helper/version.rb +3 -0
- data/spec/factories.rb +35 -0
- data/spec/factories/jobs/tmp.html +42 -0
- data/spec/factories/jobs/tmp2.html +20 -0
- data/spec/factories/jobs/tmp3.html +34 -0
- data/spec/factories/jobs_with_description.rb +20 -0
- data/spec/factories/jobs_with_title.rb +72 -0
- data/spec/preprocessors/simple_spec.rb +138 -0
- data/spec/preprocessors/with_industry_map_spec.rb +16 -0
- data/spec/selectors/n_gram_spec.rb +21 -0
- data/spec/selectors/simple_spec.rb +121 -0
- data/spec/selectors/with_binary_encoding_spec.rb +39 -0
- data/spec/spec_helper.rb +14 -0
- data/spec/support/preprocessor_spec.rb +21 -0
- data/spec/support/selector_spec.rb +21 -0
- data/svm_helper.gemspec +21 -0
- metadata +112 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: bbfc13983f715f2f0ab66d2a9dbfed543896329e
|
4
|
+
data.tar.gz: 2e6f5c32898cc01a2468ebaa2e82470e01012586
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: 8c616e62ff4717808e0b9f29e3d6773c03096934194a1fb706e3eb0ded7527c455ea00039e0a4d39fc08e13f735d4f49cd984d26128e624fbe0b956851893c21
|
7
|
+
data.tar.gz: f28756b0f3f9539f69bcda6ce3f79bdf3ef226ee813f0e5ed7b6f38595f1760e05bbd1abe7ec8ccc5c3a8db8b666871a0d33a755c19c4842a85a600e4749498d
|
data/.gitignore
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
*.gem
|
2
|
+
*.rbc
|
3
|
+
.bundle
|
4
|
+
.config
|
5
|
+
.yardoc
|
6
|
+
Gemfile.lock
|
7
|
+
InstalledFiles
|
8
|
+
_yardoc
|
9
|
+
coverage
|
10
|
+
doc/
|
11
|
+
lib/bundler/man
|
12
|
+
pkg
|
13
|
+
rdoc
|
14
|
+
spec/reports
|
15
|
+
test/tmp
|
16
|
+
test/version_tmp
|
17
|
+
tmp
|
18
|
+
|
19
|
+
.rbx
|
20
|
+
|
21
|
+
.tags
|
22
|
+
.tags_sorted_by_file
|
data/.rspec
ADDED
data/.ruby-version
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
2.0.0
|
data/.travis.yml
ADDED
data/.versions.conf
ADDED
data/.yardopts
ADDED
data/Gemfile
ADDED
@@ -0,0 +1,24 @@
|
|
1
|
+
source 'https://rubygems.org'
|
2
|
+
|
3
|
+
# Specify your gem's dependencies in svm_helper.gemspec
|
4
|
+
gemspec
|
5
|
+
|
6
|
+
group :development do
|
7
|
+
gem 'yard'
|
8
|
+
gem 'kramdown'
|
9
|
+
gem 'github-markup'
|
10
|
+
|
11
|
+
gem 'pry'
|
12
|
+
gem 'guard-rspec'
|
13
|
+
gem 'guard-yard'
|
14
|
+
|
15
|
+
gem 'rb-inotify', '~> 0.9', :require => false
|
16
|
+
gem 'rb-fsevent', :require => false
|
17
|
+
gem 'rb-fchange', :require => false
|
18
|
+
end
|
19
|
+
|
20
|
+
group :test do
|
21
|
+
gem 'rake'
|
22
|
+
gem 'mocha', require: 'mocha/api'
|
23
|
+
gem 'factory_girl', '~> 4.0'
|
24
|
+
end
|
data/Guardfile
ADDED
@@ -0,0 +1,17 @@
|
|
1
|
+
# guard 'rspec', cli: "--color --format p", all_after_pass: false do
|
2
|
+
guard 'rspec', cli: "--color --format p", all_after_pass: false, rvm:['2.0.0@svm_helper', 'jruby@svm_helper'] do
|
3
|
+
watch(%r{^spec/.+_spec\.rb$})
|
4
|
+
watch(%r{^lib/svm_helper/(.+)\.rb$}) { |m| "spec/#{m[1]}_spec.rb" }
|
5
|
+
watch('spec/spec_helper.rb') { 'spec' }
|
6
|
+
watch('spec/factories.rb') { 'spec' }
|
7
|
+
watch(%r{^spec/factories/(.+)\.rb}) { 'spec' }
|
8
|
+
watch(%r{^spec/support/(.+)_spec\.rb}) { |m| "spec/#{m[1]}s/*" }
|
9
|
+
end
|
10
|
+
|
11
|
+
notification :tmux,
|
12
|
+
:display_message => true,
|
13
|
+
:timeout => 3 # in seconds
|
14
|
+
|
15
|
+
# guard 'yard' do
|
16
|
+
# watch(%r{lib/.+\.rb})
|
17
|
+
# end
|
data/LICENSE.txt
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
Copyright (c) 2013 Andreas Eger
|
2
|
+
|
3
|
+
MIT License
|
4
|
+
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
6
|
+
a copy of this software and associated documentation files (the
|
7
|
+
"Software"), to deal in the Software without restriction, including
|
8
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
9
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
10
|
+
permit persons to whom the Software is furnished to do so, subject to
|
11
|
+
the following conditions:
|
12
|
+
|
13
|
+
The above copyright notice and this permission notice shall be
|
14
|
+
included in all copies or substantial portions of the Software.
|
15
|
+
|
16
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
17
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
18
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
19
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
20
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
21
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
22
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/README.md
ADDED
@@ -0,0 +1,41 @@
|
|
1
|
+
# SvmHelper
|
2
|
+
|
3
|
+
Shared helper classes for usage in context of SVM at experteer
|
4
|
+
|
5
|
+
[](https://travis-ci.org/sch1zo/svm_helper)
|
6
|
+
|
7
|
+
## Installation
|
8
|
+
|
9
|
+
Add this line to your application's Gemfile:
|
10
|
+
|
11
|
+
gem 'svm_helper'
|
12
|
+
|
13
|
+
And then execute:
|
14
|
+
|
15
|
+
$ bundle
|
16
|
+
|
17
|
+
Or install it yourself as:
|
18
|
+
|
19
|
+
$ gem install svm_helper
|
20
|
+
|
21
|
+
## Usage
|
22
|
+
|
23
|
+
Dataflow is normally something like this:
|
24
|
+
|
25
|
+
Job --Preprocessor--> PreprocessedData --Selector--> FeatureVector
|
26
|
+
|
27
|
+
The FeatureVector can now be used for training or prediction in a (libsvm) SVM.
|
28
|
+
|
29
|
+
Be aware that a FeatureVector has two Attributes:
|
30
|
+
|
31
|
+
data: the feature array itself
|
32
|
+
label: 1 for true, 0 for false
|
33
|
+
|
34
|
+
|
35
|
+
## Contributing
|
36
|
+
|
37
|
+
1. Fork it
|
38
|
+
2. Create your feature branch (`git checkout -b my-new-feature`)
|
39
|
+
3. Commit your changes (`git commit -am 'Add some feature'`)
|
40
|
+
4. Push to the branch (`git push origin my-new-feature`)
|
41
|
+
5. Create new Pull Request
|
data/Rakefile
ADDED
data/lib/svm_helper.rb
ADDED
@@ -0,0 +1,17 @@
|
|
1
|
+
require_relative 'interface_helper'
|
2
|
+
#
|
3
|
+
# FeatureVector interface
|
4
|
+
#
|
5
|
+
# @author Andreas Eger
|
6
|
+
class FeatureVector < InterfaceHelper
|
7
|
+
attribute :word_data
|
8
|
+
attribute :classification_arrays
|
9
|
+
attribute :labels
|
10
|
+
|
11
|
+
def label
|
12
|
+
labels[classification]
|
13
|
+
end
|
14
|
+
def data
|
15
|
+
word_data + classification_arrays[classification]
|
16
|
+
end
|
17
|
+
end
|
@@ -0,0 +1,57 @@
|
|
1
|
+
#
|
2
|
+
# @abstract Subclass and define attributes
|
3
|
+
#
|
4
|
+
# @author Andreas Eger
|
5
|
+
class InterfaceHelper
|
6
|
+
@@_attributes = Hash.new { |hash, key| hash[key] = [] }
|
7
|
+
|
8
|
+
#
|
9
|
+
# creates setter/getter similar to attr_accesor
|
10
|
+
# @param name [Symbol]
|
11
|
+
# @macro [attach] attribute
|
12
|
+
# @method $1
|
13
|
+
# reads $1
|
14
|
+
# @method $1=
|
15
|
+
# saves $1
|
16
|
+
def self.attribute name
|
17
|
+
define_method(name) do
|
18
|
+
@_attributes[name]
|
19
|
+
end
|
20
|
+
define_method(:"#{name}=") do |v|
|
21
|
+
@_attributes[name] = v
|
22
|
+
end
|
23
|
+
attributes << name unless attributes.include? name
|
24
|
+
end
|
25
|
+
def self.attributes
|
26
|
+
@@_attributes[self]
|
27
|
+
end
|
28
|
+
def initialize(params={})
|
29
|
+
@_attributes = {}
|
30
|
+
params.each do |key, value|
|
31
|
+
send("#{key}=", value)
|
32
|
+
end
|
33
|
+
@_attributes[:classification] ||= :function
|
34
|
+
end
|
35
|
+
|
36
|
+
#
|
37
|
+
# custom comperator
|
38
|
+
# @param anOther [InterfaceHelper]
|
39
|
+
#
|
40
|
+
# @return [Boolean] result after comparing each attribute
|
41
|
+
def == anOther
|
42
|
+
@_attributes.keys.map{ |sym| self.send(sym) == anOther.send(sym)}.reduce(true){|a,e| a && e }
|
43
|
+
end
|
44
|
+
|
45
|
+
def industry!
|
46
|
+
@_attributes[:classification] = :industry
|
47
|
+
end
|
48
|
+
def function!
|
49
|
+
@_attributes[:classification] = :function
|
50
|
+
end
|
51
|
+
def career_level!
|
52
|
+
@_attributes[:classification] = :career_level
|
53
|
+
end
|
54
|
+
def classification
|
55
|
+
@_attributes[:classification]
|
56
|
+
end
|
57
|
+
end
|
@@ -0,0 +1,17 @@
|
|
1
|
+
require_relative 'interface_helper'
|
2
|
+
#
|
3
|
+
# PreprocessedData interface
|
4
|
+
#
|
5
|
+
# @author Andreas Eger
|
6
|
+
class PreprocessedData < InterfaceHelper
|
7
|
+
attribute :data
|
8
|
+
attribute :ids
|
9
|
+
attribute :labels
|
10
|
+
|
11
|
+
def id
|
12
|
+
ids[classification]
|
13
|
+
end
|
14
|
+
def label
|
15
|
+
labels[classification]
|
16
|
+
end
|
17
|
+
end
|
@@ -0,0 +1,111 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
module Preprocessor
|
3
|
+
#
|
4
|
+
# Preprocessor which just cleans to text
|
5
|
+
#
|
6
|
+
# @author Andreas Eger
|
7
|
+
#
|
8
|
+
class Simple
|
9
|
+
THREAD_COUNT = (ENV['OMP_NUM_THREADS'] || 2).to_i
|
10
|
+
# filters most gender stuff
|
11
|
+
GENDER_FILTER = %r{(\(*(m|w)(\/|\|)(w|m)\)*)|(/-*in)|\(in\)}
|
12
|
+
# filters most wierd symbols
|
13
|
+
SYMBOL_FILTER = %r{/|-|–|:|\+|!|,|\.|\*|\?|/|·|\"|„|•||\||(\S*(&|;)\S*)}
|
14
|
+
# urls and email filter
|
15
|
+
URL_FILTER = /(https?:\/\/)?([\da-z\.-]+)\.([a-z\.]{2,6})([\/\w \.-]*)*\/?/
|
16
|
+
EMAIL_FILTER = /([a-z0-9_\.-]+)@([\da-z\.-]+)\.([a-z\.]{2,6})/
|
17
|
+
# filter for new lines
|
18
|
+
NEW_LINES = /(\r\n)|\r|\n/
|
19
|
+
# extract words from brackets
|
20
|
+
WORDS_IN_BRACKETS = /\(([a-zA-Z]+)\)/
|
21
|
+
# filters multiple whitesspace
|
22
|
+
WHITESPACE = /(\s| )+/
|
23
|
+
# filters all kind of XMl/HTML tags
|
24
|
+
XML_TAG_FILTER = /<(.*?)>/
|
25
|
+
# filter for used job tokens
|
26
|
+
CODE_TOKEN_FILTER = /\[[^\]]*\]|\([^\)]*\)|\{[^\}]*\}|\S*\d+\w+/
|
27
|
+
|
28
|
+
def initialize args={}
|
29
|
+
@parallel = args.fetch(:parallel){false}
|
30
|
+
end
|
31
|
+
|
32
|
+
def label
|
33
|
+
"simple"
|
34
|
+
end
|
35
|
+
#
|
36
|
+
# cleans provided jobs
|
37
|
+
# @overload process(jobs, classification)
|
38
|
+
# @param jobs [Job] single Job
|
39
|
+
# @param classification [Symbol] in `:industry`, `:function`, `:career_level`
|
40
|
+
# @overload process(jobs, classification)
|
41
|
+
# @param jobs [Array<Job>] list of Jobs
|
42
|
+
# @param classification [Symbol] in `:industry`, `:function`, `:career_level`
|
43
|
+
#
|
44
|
+
# @return [Array<PreprocessedData>] list of processed job data - or singe job data
|
45
|
+
def process jobs, classification=:function
|
46
|
+
if jobs.respond_to? :map
|
47
|
+
process_jobs jobs, classification
|
48
|
+
else
|
49
|
+
process_job jobs, classification
|
50
|
+
end
|
51
|
+
end
|
52
|
+
|
53
|
+
#
|
54
|
+
# converts string into a cleaner version
|
55
|
+
# @param title [String] job title
|
56
|
+
#
|
57
|
+
# @return [String] clean and lowercase version of input
|
58
|
+
def clean_title title
|
59
|
+
title.gsub(GENDER_FILTER,'').
|
60
|
+
gsub(SYMBOL_FILTER,'').
|
61
|
+
gsub(WORDS_IN_BRACKETS, '\1').
|
62
|
+
gsub(CODE_TOKEN_FILTER,'').
|
63
|
+
gsub(WHITESPACE,' ').
|
64
|
+
downcase.
|
65
|
+
strip
|
66
|
+
end
|
67
|
+
#
|
68
|
+
# converts string into a cleaner version
|
69
|
+
# @param desc [String] job description
|
70
|
+
#
|
71
|
+
# @return [String] clean and lowercase version of input
|
72
|
+
def clean_description desc
|
73
|
+
desc.gsub(XML_TAG_FILTER,' ')
|
74
|
+
.gsub(EMAIL_FILTER,'')
|
75
|
+
.gsub(URL_FILTER,'')
|
76
|
+
.gsub(GENDER_FILTER,'')
|
77
|
+
.gsub(NEW_LINES,'')
|
78
|
+
.gsub(SYMBOL_FILTER,' ')
|
79
|
+
.gsub(WHITESPACE,' ')
|
80
|
+
.gsub(WORDS_IN_BRACKETS, '\1')
|
81
|
+
.gsub(CODE_TOKEN_FILTER,'')
|
82
|
+
.downcase
|
83
|
+
.strip
|
84
|
+
end
|
85
|
+
|
86
|
+
private
|
87
|
+
def process_jobs jobs, classification
|
88
|
+
if @parallel && RUBY_PLATFORM == 'java'
|
89
|
+
Parallel.map(jobs, in_threads: THREAD_COUNT ) {|job| process_job job, classification }
|
90
|
+
elsif @parallel
|
91
|
+
Parallel.map(jobs, in_processes: THREAD_COUNT ) {|job| process_job job, classification }
|
92
|
+
else
|
93
|
+
jobs.map {|job| process_job job, classification }
|
94
|
+
end
|
95
|
+
end
|
96
|
+
|
97
|
+
def process_job job, classification
|
98
|
+
PreprocessedData.new(
|
99
|
+
data: [ clean_title(job.title), clean_description(job.description) ],
|
100
|
+
ids: {
|
101
|
+
industry: job.classification_id(:industry),
|
102
|
+
function: job.classification_id(:function),
|
103
|
+
career_level: job.classification_id(:career_level) },
|
104
|
+
labels: {
|
105
|
+
industry: job.label(:industry),
|
106
|
+
function: job.label(:function),
|
107
|
+
career_level: job.label(:career_level) }
|
108
|
+
).tap{|e| e.send("#{classification}!")}
|
109
|
+
end
|
110
|
+
end
|
111
|
+
end
|
@@ -0,0 +1,40 @@
|
|
1
|
+
require_relative 'simple'
|
2
|
+
module Preprocessor
|
3
|
+
#
|
4
|
+
# Preprocessor Base Class
|
5
|
+
#
|
6
|
+
# @author Andreas Eger
|
7
|
+
#
|
8
|
+
class WithIndustryMap < Simple
|
9
|
+
attr_reader :industry_map
|
10
|
+
|
11
|
+
#
|
12
|
+
# @param args [Hash] options hash
|
13
|
+
# @option args [Hash] :industry_map mapping for the tree like industry ids to continuous ones
|
14
|
+
def initialize args={}
|
15
|
+
@industry_map = args.fetch(:industry_map){ Hash[Pjpp::Industry.select(:id).all.map(&:id).sort.map.with_index{|e,i| [e,i]}] }
|
16
|
+
end
|
17
|
+
|
18
|
+
def map_industry_id(id)
|
19
|
+
@industry_map[id]
|
20
|
+
end
|
21
|
+
def label
|
22
|
+
"with_industry_map"
|
23
|
+
end
|
24
|
+
|
25
|
+
private
|
26
|
+
def process_job job, classification
|
27
|
+
PreprocessedData.new(
|
28
|
+
data: [ clean_title(job.title), clean_description(job.description) ],
|
29
|
+
ids: {
|
30
|
+
industry: map_industry_id(job.classification_id(:industry)),
|
31
|
+
function: job.classification_id(:function),
|
32
|
+
career_level: job.classification_id(:career_level) },
|
33
|
+
labels: {
|
34
|
+
industry: job.label(:industry),
|
35
|
+
function: job.label(:function),
|
36
|
+
career_level: job.label(:career_level) }
|
37
|
+
).tap{|e| e.send("#{classification}!")}
|
38
|
+
end
|
39
|
+
end
|
40
|
+
end
|