simple_similarity 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +17 -0
- data/Gemfile +4 -0
- data/LICENSE.txt +22 -0
- data/README.md +79 -0
- data/Rakefile +1 -0
- data/lib/simple_similarity.rb +78 -0
- data/lib/simple_similarity/version.rb +3 -0
- data/simple_similarity.gemspec +23 -0
- data/spec/benchmark.rb +18 -0
- data/spec/simple_similarity_spec.rb +41 -0
- metadata +122 -0
data/.gitignore
ADDED
data/Gemfile
ADDED
data/LICENSE.txt
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
Copyright (c) 2013 Ori Pekelman
|
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 +1,79 @@
|
|
1
|
+
# SimpleSimilarity
|
2
|
+
|
3
|
+
This is an implemenation of a very trivial, and very costly algorithm to find
|
4
|
+
Similarity in a collection of items. It calculates a distance between the
|
5
|
+
length of the items compressed each on its own, and compressed when concatenated
|
6
|
+
The theory is sound.
|
7
|
+
|
8
|
+
## Installation
|
9
|
+
|
10
|
+
Add this line to your application's Gemfile:
|
11
|
+
|
12
|
+
gem 'simple_similarity'
|
13
|
+
|
14
|
+
And then execute:
|
15
|
+
|
16
|
+
$ bundle
|
17
|
+
|
18
|
+
Or install it yourself as:
|
19
|
+
|
20
|
+
$ gem install simple_similarity
|
21
|
+
|
22
|
+
## Usage
|
23
|
+
|
24
|
+
Get the distance between "Hello" and "World", always takes to arguments.
|
25
|
+
|
26
|
+
SimpleSimilarity::SimpleSimilarity.distance("Hello","World")
|
27
|
+
|
28
|
+
$ 0.125
|
29
|
+
|
30
|
+
SimpleSimilarity::SimpleSimilarity.case_insensitive_distance("Hello","hello")
|
31
|
+
|
32
|
+
$ 1.0
|
33
|
+
|
34
|
+
Get a distance matrix, the first argument is an array of elements, the second
|
35
|
+
is a cutoff ratio for the detected similarities (0.5 is quite similar, above
|
36
|
+
0.8 we are basically identical, depending on the length of the original strings..)
|
37
|
+
|
38
|
+
SimpleSimilarity::SimpleSimilarity.distance_matrix(%w( hello, hell, world), 0.5, true, false)
|
39
|
+
|
40
|
+
$ => {
|
41
|
+
:matrix=>[
|
42
|
+
[nil, 0.5833333333333333, 0.11764705882352944],
|
43
|
+
[nil, nil, 0.125],
|
44
|
+
[nil, nil, nil]
|
45
|
+
],
|
46
|
+
:similar_terms=>[{
|
47
|
+
:first_term=>[0, "hello,"],
|
48
|
+
:second_term=>[1, "hell,"],
|
49
|
+
:score=>0.5833333333333333}]}
|
50
|
+
|
51
|
+
### Caveat :
|
52
|
+
1. this algorithm costs (n^2/2) - n in time and n^2 in memory
|
53
|
+
>So if n= 100,000
|
54
|
+
> there will be 4,999,900,000 calculations
|
55
|
+
2. setting only_hits to true does not return the matrix only the hits but optimizes memory usage.
|
56
|
+
### Note:
|
57
|
+
the matrix can be used later for clustering....
|
58
|
+
|
59
|
+
#Testing
|
60
|
+
spec:
|
61
|
+
|
62
|
+
bundle exec rspec spec
|
63
|
+
|
64
|
+
|
65
|
+
Benchmark:
|
66
|
+
|
67
|
+
On my machine comparing 100 items takes around 8 seconds:
|
68
|
+
|
69
|
+
bundle exec ruby spec/benchmark.rb
|
70
|
+
|
71
|
+
If what you want to compare are longer documents, swap the
|
72
|
+
|
73
|
+
## Contributing
|
74
|
+
|
75
|
+
1. Fork it
|
76
|
+
2. Create your feature branch (`git checkout -b my-new-feature`)
|
77
|
+
3. Commit your changes (`git commit -am 'Add some feature'`)
|
78
|
+
4. Push to the branch (`git push origin my-new-feature`)
|
79
|
+
5. Create new Pull Request
|
data/Rakefile
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
require "bundler/gem_tasks"
|
@@ -0,0 +1,78 @@
|
|
1
|
+
require "simple_similarity/version"
|
2
|
+
require "zlib"
|
3
|
+
|
4
|
+
module SimpleSimilarity
|
5
|
+
class SimpleSimilarity
|
6
|
+
#calculate similarity by comparing the length of compressed(a) + length of
|
7
|
+
#compressed (b) to the lengt of compressed (a and b concatenated)
|
8
|
+
def self.distance(a, b)
|
9
|
+
a = a.to_s
|
10
|
+
b = b.to_s
|
11
|
+
|
12
|
+
# optimization: well if they are equal lets not do expensive stuff
|
13
|
+
if a==b then return 1.0 end
|
14
|
+
|
15
|
+
# optimization: if one is empty and the other not they are totally different
|
16
|
+
if ((a.empty? && !b.empty?) || (b.empty? && !a.empty?)) then return (0.to_f) end
|
17
|
+
|
18
|
+
# This is hacky, but useful. When strings are too short the compression
|
19
|
+
# is not sufficient.
|
20
|
+
if (a.length < 40 || b.length < 40)
|
21
|
+
a = a * 10
|
22
|
+
b = b * 10
|
23
|
+
end
|
24
|
+
|
25
|
+
# 8 characters is the minimal zlib compressed stream, this is basically a constant length we remove
|
26
|
+
minimal_zip_length = 7
|
27
|
+
azip = Zlib.deflate(a).length - minimal_zip_length
|
28
|
+
bzip = Zlib.deflate(b).length - minimal_zip_length
|
29
|
+
|
30
|
+
aplusbzip = (Zlib.deflate(a + b).length) - minimal_zip_length
|
31
|
+
((azip + bzip).to_f/aplusbzip.to_f) - 1 # -1 is because something that is totally different will get 1.00 totally the same 2.00.
|
32
|
+
end
|
33
|
+
|
34
|
+
#Normalize the string, downcase remove non alphanumeric characters
|
35
|
+
def self.case_insensitive_distance(a, b)
|
36
|
+
self.distance(a.downcase.gsub(/[^[:alnum:]]/, ' '), b.downcase.gsub(/[^[:alnum:]]/, ' '))
|
37
|
+
end
|
38
|
+
|
39
|
+
# This method will return a distance matrix as well as matches beyond a supplies threshold
|
40
|
+
# Caveat : this algorithm costs (n^2/2) - n in time and n^2 in memory
|
41
|
+
# So if n= 100,000
|
42
|
+
# there will be 4,999,900,000 calculations
|
43
|
+
# setting only_hits to true does not return the matrix only the hits
|
44
|
+
# the matrix can be used later for clustering....
|
45
|
+
|
46
|
+
# "hello", "hello", "world", "hello world", "hello cruel world"
|
47
|
+
# "hello" [nil, 1.0, 0.125, 0.41176470588235303, 0.28],
|
48
|
+
# "hello" [nil, nil, 0.125, 0.41176470588235303, 0.28],
|
49
|
+
# "world" [nil, nil, nil, 0.41176470588235303, 0.28],
|
50
|
+
# "hello world" [nil, nil, nil, nil, 0.52],
|
51
|
+
# "hello cruel world" [nil, nil, nil, nil, nil]
|
52
|
+
def self.distance_matrix(arr, cutoff = 0.5, case_insensitive = false, only_hits = true )
|
53
|
+
sims =[]
|
54
|
+
len = arr.length
|
55
|
+
if only_hits
|
56
|
+
mat = []
|
57
|
+
else
|
58
|
+
mat = Array.new(len){Array.new(len)}
|
59
|
+
end
|
60
|
+
|
61
|
+
dist = 0.to_f
|
62
|
+
for i in 0..len
|
63
|
+
for j in i+1..len -1
|
64
|
+
if case_insensitive then
|
65
|
+
dist = self.case_insensitive_distance(arr[i], arr[j])
|
66
|
+
else
|
67
|
+
dist = self.distance(arr[i], arr[j])
|
68
|
+
end
|
69
|
+
if dist >= cutoff
|
70
|
+
sims.push({:first_term=>[i,arr[i]],:second_term=>[j,arr[j]],:score=>dist})
|
71
|
+
end
|
72
|
+
mat[i][j] = dist unless only_hits
|
73
|
+
end
|
74
|
+
end
|
75
|
+
{:matrix=>mat,:similar_terms=>sims}
|
76
|
+
end
|
77
|
+
end
|
78
|
+
end
|
@@ -0,0 +1,23 @@
|
|
1
|
+
# -*- encoding: utf-8 -*-
|
2
|
+
lib = File.expand_path('../lib', __FILE__)
|
3
|
+
$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
|
4
|
+
require 'simple_similarity/version'
|
5
|
+
|
6
|
+
Gem::Specification.new do |gem|
|
7
|
+
gem.name = "simple_similarity"
|
8
|
+
gem.version = SimpleSimilarity::VERSION
|
9
|
+
gem.authors = ["Ori Pekelman"]
|
10
|
+
gem.email = ["ori@pekelman.com"]
|
11
|
+
gem.description = %q{A simple (slow but beautiful) algorithm to find similar things}
|
12
|
+
gem.summary = %q{A simple (slow but beautiful) algorithm to find similar things}
|
13
|
+
gem.homepage = "https://github.com/OriPekelman/simple_similarity"
|
14
|
+
|
15
|
+
gem.files = `git ls-files`.split($/)
|
16
|
+
gem.executables = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
|
17
|
+
gem.test_files = gem.files.grep(%r{^(test|spec|features)/})
|
18
|
+
gem.require_paths = ["lib"]
|
19
|
+
gem.add_development_dependency "rspec", "~> 2.6"
|
20
|
+
gem.add_development_dependency "pry"
|
21
|
+
gem.add_development_dependency "awesome_print"
|
22
|
+
gem.add_development_dependency "colorize"
|
23
|
+
end
|
data/spec/benchmark.rb
ADDED
@@ -0,0 +1,18 @@
|
|
1
|
+
require 'simple_similarity'
|
2
|
+
require "benchmark"
|
3
|
+
|
4
|
+
a = []
|
5
|
+
CHARS = (?a..?z).to_a + ["e","e","e","e","e","a","a","a","a","a","a","a","a","a","a","a","a"]
|
6
|
+
1000.times {a.push(5.times.inject("") {|s, i| s << CHARS[rand(CHARS.size)]})}
|
7
|
+
|
8
|
+
time = Benchmark.measure do
|
9
|
+
puts SimpleSimilarity::SimpleSimilarity.distance_matrix(a, 0.5, false, true)[:similar_terms]
|
10
|
+
end
|
11
|
+
puts "Without matrix"
|
12
|
+
puts time
|
13
|
+
|
14
|
+
time = Benchmark.measure do
|
15
|
+
puts SimpleSimilarity::SimpleSimilarity.distance_matrix(a, 0.5, false, false)[:similar_terms]
|
16
|
+
end
|
17
|
+
puts "With matrix"
|
18
|
+
puts time
|
@@ -0,0 +1,41 @@
|
|
1
|
+
require 'simple_similarity'
|
2
|
+
|
3
|
+
describe SimpleSimilarity::SimpleSimilarity do
|
4
|
+
it "Hello is identical to Hello" do
|
5
|
+
SimpleSimilarity::SimpleSimilarity.distance("Hello","Hello").should eql(1.0)
|
6
|
+
end
|
7
|
+
|
8
|
+
it "Hello is not similar to world" do
|
9
|
+
SimpleSimilarity::SimpleSimilarity.distance("Hello","World").should eql(0.125)
|
10
|
+
end
|
11
|
+
|
12
|
+
it "Hello is somewhat similar hello" do
|
13
|
+
SimpleSimilarity::SimpleSimilarity.distance("Hello","hello").should eql(0.5)
|
14
|
+
end
|
15
|
+
|
16
|
+
it "Hello is identical to hello when case insensitive" do
|
17
|
+
SimpleSimilarity::SimpleSimilarity.case_insensitive_distance("Hello","hello").should eql(1.0)
|
18
|
+
end
|
19
|
+
|
20
|
+
it "from : hello, Hello, hell, world, cruel, crude with a cutoff of 0.5 we should have two couples" do
|
21
|
+
SimpleSimilarity::SimpleSimilarity.distance_matrix(%w( hello, Hello, hell, world, cruel, crud), 0.5)[:similar_terms].length.should eql(2)
|
22
|
+
end
|
23
|
+
|
24
|
+
it "from : hello, Hello, hell, world, cruel, crude with a cutoff of 0.5 we should have three couples case insensitive" do
|
25
|
+
SimpleSimilarity::SimpleSimilarity.distance_matrix(%w( hello, Hello, hell, world, cruel, crud), 0.5, true)[:similar_terms].length.should eql(3)
|
26
|
+
end
|
27
|
+
|
28
|
+
it "from : hello world, Hello world, hell world, world, cruel, crude with a cutoff of 1.0 we should have seven couples case insensitive" do
|
29
|
+
SimpleSimilarity::SimpleSimilarity.distance_matrix(%w( hello world, Hello world, hell world, world, cruel, crude), 1.0, true)[:similar_terms].length.should eql(7)
|
30
|
+
end
|
31
|
+
|
32
|
+
it "setting only_hits to false gives back a matrix" do
|
33
|
+
SimpleSimilarity::SimpleSimilarity.distance_matrix(%w( hello, world), 1.0, true, false)[:matrix].length.should eql(2)
|
34
|
+
end
|
35
|
+
|
36
|
+
it "by default does not gives an empty matrix" do
|
37
|
+
SimpleSimilarity::SimpleSimilarity.distance_matrix(%w( hello, world))[:matrix].length.should eql(0)
|
38
|
+
end
|
39
|
+
|
40
|
+
|
41
|
+
end
|
metadata
ADDED
@@ -0,0 +1,122 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: simple_similarity
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
prerelease:
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Ori Pekelman
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
date: 2013-01-08 00:00:00.000000000 Z
|
13
|
+
dependencies:
|
14
|
+
- !ruby/object:Gem::Dependency
|
15
|
+
name: rspec
|
16
|
+
requirement: !ruby/object:Gem::Requirement
|
17
|
+
none: false
|
18
|
+
requirements:
|
19
|
+
- - ~>
|
20
|
+
- !ruby/object:Gem::Version
|
21
|
+
version: '2.6'
|
22
|
+
type: :development
|
23
|
+
prerelease: false
|
24
|
+
version_requirements: !ruby/object:Gem::Requirement
|
25
|
+
none: false
|
26
|
+
requirements:
|
27
|
+
- - ~>
|
28
|
+
- !ruby/object:Gem::Version
|
29
|
+
version: '2.6'
|
30
|
+
- !ruby/object:Gem::Dependency
|
31
|
+
name: pry
|
32
|
+
requirement: !ruby/object:Gem::Requirement
|
33
|
+
none: false
|
34
|
+
requirements:
|
35
|
+
- - ! '>='
|
36
|
+
- !ruby/object:Gem::Version
|
37
|
+
version: '0'
|
38
|
+
type: :development
|
39
|
+
prerelease: false
|
40
|
+
version_requirements: !ruby/object:Gem::Requirement
|
41
|
+
none: false
|
42
|
+
requirements:
|
43
|
+
- - ! '>='
|
44
|
+
- !ruby/object:Gem::Version
|
45
|
+
version: '0'
|
46
|
+
- !ruby/object:Gem::Dependency
|
47
|
+
name: awesome_print
|
48
|
+
requirement: !ruby/object:Gem::Requirement
|
49
|
+
none: false
|
50
|
+
requirements:
|
51
|
+
- - ! '>='
|
52
|
+
- !ruby/object:Gem::Version
|
53
|
+
version: '0'
|
54
|
+
type: :development
|
55
|
+
prerelease: false
|
56
|
+
version_requirements: !ruby/object:Gem::Requirement
|
57
|
+
none: false
|
58
|
+
requirements:
|
59
|
+
- - ! '>='
|
60
|
+
- !ruby/object:Gem::Version
|
61
|
+
version: '0'
|
62
|
+
- !ruby/object:Gem::Dependency
|
63
|
+
name: colorize
|
64
|
+
requirement: !ruby/object:Gem::Requirement
|
65
|
+
none: false
|
66
|
+
requirements:
|
67
|
+
- - ! '>='
|
68
|
+
- !ruby/object:Gem::Version
|
69
|
+
version: '0'
|
70
|
+
type: :development
|
71
|
+
prerelease: false
|
72
|
+
version_requirements: !ruby/object:Gem::Requirement
|
73
|
+
none: false
|
74
|
+
requirements:
|
75
|
+
- - ! '>='
|
76
|
+
- !ruby/object:Gem::Version
|
77
|
+
version: '0'
|
78
|
+
description: A simple (slow but beautiful) algorithm to find similar things
|
79
|
+
email:
|
80
|
+
- ori@pekelman.com
|
81
|
+
executables: []
|
82
|
+
extensions: []
|
83
|
+
extra_rdoc_files: []
|
84
|
+
files:
|
85
|
+
- .gitignore
|
86
|
+
- Gemfile
|
87
|
+
- LICENSE.txt
|
88
|
+
- README.md
|
89
|
+
- Rakefile
|
90
|
+
- lib/simple_similarity.rb
|
91
|
+
- lib/simple_similarity/version.rb
|
92
|
+
- simple_similarity.gemspec
|
93
|
+
- spec/benchmark.rb
|
94
|
+
- spec/simple_similarity_spec.rb
|
95
|
+
homepage: https://github.com/OriPekelman/simple_similarity
|
96
|
+
licenses: []
|
97
|
+
post_install_message:
|
98
|
+
rdoc_options: []
|
99
|
+
require_paths:
|
100
|
+
- lib
|
101
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
102
|
+
none: false
|
103
|
+
requirements:
|
104
|
+
- - ! '>='
|
105
|
+
- !ruby/object:Gem::Version
|
106
|
+
version: '0'
|
107
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
108
|
+
none: false
|
109
|
+
requirements:
|
110
|
+
- - ! '>='
|
111
|
+
- !ruby/object:Gem::Version
|
112
|
+
version: '0'
|
113
|
+
requirements: []
|
114
|
+
rubyforge_project:
|
115
|
+
rubygems_version: 1.8.24
|
116
|
+
signing_key:
|
117
|
+
specification_version: 3
|
118
|
+
summary: A simple (slow but beautiful) algorithm to find similar things
|
119
|
+
test_files:
|
120
|
+
- spec/benchmark.rb
|
121
|
+
- spec/simple_similarity_spec.rb
|
122
|
+
has_rdoc:
|