hits 0.0.1

data/LICENSE

@@ -0,0 +1,3 @@
+== hits
+
+Put appropriate LICENSE for your project here.

data/README

@@ -0,0 +1,42 @@
+== hits
+
+A poor man's implementation of Jon Kleinberg's Hyperlink-Induced Topic Search (HITS) (also known as Hubs and authorities). See http://en.wikipedia.org/wiki/HITS_algorithm
+
+require 'rubygems'
+require 'hits'
+
+# create a graph
+graph = Hits::Graph.new
+
+# add some edges to the graph with weights
+graph.add_edge(:bsbodden, :objo, 1.0)
+graph.add_edge(:bsbodden, :nusairat, 2.0)
+graph.add_edge(:bsbodden, :looselytyped, 3.0)
+graph.add_edge(:bsbodden, :neal4d, 1.5)
+graph.add_edge(:objo, :nusairat, 2.5)
+graph.add_edge(:objo, :bsbodden, 1.0)
+graph.add_edge(:neal4d, :bsbodden, 1.15)
+graph.add_edge(:nusairat, :bsbodden, 4.5)
+
+# textual display of the graph
+puts "graph ==> #{graph}"
+
+# create a HITS for the graph
+hits = Hits::Hits.new(graph)
+
+# show the vertexes incoming and outgoing links (inlinks and outlinks)
+graph.each_vertex { |v| puts "in links for #{v} ==> #{graph.in_links(v)}, out links for #{v} ==> #{graph.out_links(v)}"}
+
+# compute HITS with the default number of iterations
+hits.compute_hits
+
+# print the top HUBS and AUTHORITIES
+puts "=== TOP HUBS ==="
+hits.top_hub_scores.each do |hub|
+  puts "hub #{hub}"
+end
+
+puts "=== TOP AUTHORITIES ==="
+hits.top_authority_scores.each do |authority|
+  puts "authority #{authority}"
+end

data/Rakefile

@@ -0,0 +1,44 @@
+require 'rubygems'
+require 'rake'
+require 'rake/clean'
+require 'rake/gempackagetask'
+require 'rake/rdoctask'
+require 'rake/testtask'
+require 'spec/rake/spectask'
+
+spec = Gem::Specification.new do |s|
+  s.name = 'hits'
+  s.version = '0.0.1'
+  s.has_rdoc = true
+  s.extra_rdoc_files = ['README', 'LICENSE']
+  s.summary = "A poor man's implementation of Jon Kleinberg's Hyperlink-Induced Topic Search (HITS) (also known as Hubs and authorities)"
+  s.description = s.summary
+  s.author = 'Brian Sam-Bodden'
+  s.email = 'bsbodden@integrallis.com'
+  s.files = %w(LICENSE README Rakefile) + Dir.glob("{bin,lib,spec}/**/*")
+  s.require_path = "lib"
+  s.bindir = "bin"
+end
+
+Rake::GemPackageTask.new(spec) do |p|
+  p.gem_spec = spec
+  p.need_tar = true
+  p.need_zip = true
+end
+
+Rake::RDocTask.new do |rdoc|
+  files =['README', 'LICENSE', 'lib/**/*.rb']
+  rdoc.rdoc_files.add(files)
+  rdoc.main = "README" # page to start on
+  rdoc.title = "hits Docs"
+  rdoc.rdoc_dir = 'doc/rdoc' # rdoc output folder
+  rdoc.options << '--line-numbers'
+end
+
+Rake::TestTask.new do |t|
+  t.test_files = FileList['test/**/*.rb']
+end
+
+Spec::Rake::SpecTask.new do |t|
+  t.spec_files = FileList['spec/**/*.rb']
+end

data/lib/hits/graph.rb

@@ -0,0 +1,48 @@
+require 'rgl/dot'
+require 'rgl/adjacency'
+require 'rgl/bidirectional'
+
+module Hits
+  
+  class Graph
+    attr_reader :graph
+    
+    def initialize
+      @graph = RGL::DirectedAdjacencyGraph.new
+      @in_links = {}
+      @edge_weights = {}
+    end
+    
+    def add_edge(from, to, weight = 1.0)
+      @graph.add_edge(from, to)
+      @in_links[to] ||= []
+      @in_links[to] << from unless @in_links[to].include? from
+      @edge_weights[[to, from]] = weight
+    end
+    
+    def in_links(vertex)
+      @in_links[vertex]
+    end
+    
+    def out_links(vertex)
+      @graph.adjacent_vertices(vertex)
+    end
+    
+    def each_vertex(&b)
+      @graph.each_vertex(&b)
+    end
+    
+    def weight(to, from)
+      @edge_weights[[to, from]]
+    end
+    
+    def weight=(to, from, weight)
+      @edge_weights[[to, from]] = weight if @edge_weights[[to, from]]
+    end
+
+    def to_s
+      @graph.edges.to_a.to_s
+    end
+      
+  end
+end

data/lib/hits/hits.rb

@@ -0,0 +1,36 @@
+module Hits
+
+  class Hits
+    
+    def initialize(graph)
+      @graph = graph
+      @hub_scores = {}
+      @authority_scores = {}
+      @graph.each_vertex do |vertex|
+        @hub_scores[vertex] = 1.0
+        @authority_scores[vertex] = 1.0
+      end
+    end
+    
+    def compute_hits(iterations = 25)
+      (1..iterations).each do
+        @graph.each_vertex do |vertex|
+          authority_score = @graph.in_links(vertex).inject(0.0) { |sum, vertex| sum + @hub_scores[vertex] } if @graph.in_links(vertex)
+          hub_score = @graph.out_links(vertex).inject(0.0) { |sum, vertex| sum + @authority_scores[vertex] } if @graph.out_links(vertex)
+          @authority_scores[vertex] = authority_score || 0.0
+          @hub_scores[vertex] = hub_score || 0.0   
+        end
+      end
+    end
+    
+    def top_hub_scores(how_many=5)
+      @hub_scores.sort_by { |k,v| v }.collect { |v| v[0] }.reverse.first(how_many)
+    end
+    
+    def top_authority_scores(how_many=5)
+      @authority_scores.sort_by { |k,v| v }.collect { |v| v[0] }.reverse.first(how_many)
+    end
+
+  end
+     
+end

data/lib/hits.rb

@@ -0,0 +1,2 @@
+require 'hits/hits'
+require 'hits/graph'

metadata

@@ -0,0 +1,61 @@
+--- !ruby/object:Gem::Specification 
+name: hits
+version: !ruby/object:Gem::Version 
+  version: 0.0.1
+platform: ruby
+authors: 
+- Brian Sam-Bodden
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2009-11-08 00:00:00 -07:00
+default_executable: 
+dependencies: []
+
+description: A poor man's implementation of Jon Kleinberg's Hyperlink-Induced Topic Search (HITS) (also known as Hubs and authorities)
+email: bsbodden@integrallis.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: 
+- README
+- LICENSE
+files: 
+- LICENSE
+- README
+- Rakefile
+- lib/hits/graph.rb
+- lib/hits/hits.rb
+- lib/hits.rb
+has_rdoc: true
+homepage: 
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.3.5
+signing_key: 
+specification_version: 3
+summary: A poor man's implementation of Jon Kleinberg's Hyperlink-Induced Topic Search (HITS) (also known as Hubs and authorities)
+test_files: []
+