diff --git a/lib/tasks/tag_clustering.rake b/lib/tasks/tag_clustering.rake
index e527c2d96c35d21a5db0a116464bb5ee048241fc..4828780a88bafbe672ab5d7069e92fc012019749 100644
--- a/lib/tasks/tag_clustering.rake
+++ b/lib/tasks/tag_clustering.rake
@@ -7,63 +7,63 @@ namespace :tag do
hash = {}
edges_total = 0
- graphPath = Rails.root.join(outDir, fileName + ".net")
-
- def swap(a, b)
- tmp = a
- a = b
- b = tmp
- end
-
- LearningObject.all.each do |lo|
- # for each lo, count tags and tag pairs and add to hash
- # if id1 <= id2
- # hash[id1][id2] will equal how many times tags with id1 and id2 appear together on a LO
- lo.tags.each.with_index do |t, i|
- hash[t.id] = {} if hash[t.id].nil?
- hash[t.id][t.id] = 0 if hash[t.id][t.id].nil?
- hash[t.id][t.id] += 1
- lo.tags.drop(i+1).each do |t2|
- if t.id > t2.id
- swap(t, t2)
- hash[t.id] = {} if hash[t.id].nil?
- end
- if hash[t.id][t2.id].nil?
- hash[t.id][t2.id] = 0
- end
- hash[t2.id] = {} if hash[t2.id].nil?
-
- # hash[t2.id][t2.id] = 0 if hash[t2.id][t2.id].nil?
- # hash[t2.id][t2.id] += 1
- # hash[t.id][t.id] = 0 if hash[t.id][t.id].nil?
- # hash[t.id][t.id] += 1
-
- hash[t.id][t2.id] += 1
- end
- end
- end
-
- File.open(graphPath, "w+") do |f|
- f << "*Vertices #{Tag.all.size}\n"
- # tags = Tag.all.to_ary
- tag_index = {}
- Tag.all.each_with_index do |t,i|
- f << "#{i+1} \"#{t.name}\"\n"
- tag_index[t.id] = i+1
- end
-
- f << "*Edges #{edges_total}\n"
-
- hash.each do |id1, ids2Hash|
- ids2Hash.each do |id2, value|
- if id1 != id2
- f << "#{tag_index[id1]} #{tag_index[id2]} #{hash[id1][id2].to_f/(Math.sqrt(hash[id1][id1]*hash[id2][id2]))}\n"
- end
- end
- end
- end
-
- system("infomap --ftree #{graphPath} #{Rails.root.join(outDIR)}")
+ graphPath = Rails.root.join(outDIR, fileName + ".net")
+ #
+ # def swap(a, b)
+ # tmp = a
+ # a = b
+ # b = tmp
+ # end
+ #
+ # LearningObject.all.each do |lo|
+ # # for each lo, count tags and tag pairs and add to hash
+ # # if id1 <= id2
+ # # hash[id1][id2] will equal how many times tags with id1 and id2 appear together on a LO
+ # lo.tags.each.with_index do |t, i|
+ # hash[t.id] = {} if hash[t.id].nil?
+ # hash[t.id][t.id] = 0 if hash[t.id][t.id].nil?
+ # hash[t.id][t.id] += 1
+ # lo.tags.drop(i+1).each do |t2|
+ # if t.id > t2.id
+ # swap(t, t2)
+ # hash[t.id] = {} if hash[t.id].nil?
+ # end
+ # if hash[t.id][t2.id].nil?
+ # hash[t.id][t2.id] = 0
+ # end
+ # hash[t2.id] = {} if hash[t2.id].nil?
+ #
+ # # hash[t2.id][t2.id] = 0 if hash[t2.id][t2.id].nil?
+ # # hash[t2.id][t2.id] += 1
+ # # hash[t.id][t.id] = 0 if hash[t.id][t.id].nil?
+ # # hash[t.id][t.id] += 1
+ #
+ # hash[t.id][t2.id] += 1
+ # end
+ # end
+ # end
+ #
+ # File.open(graphPath, "w+") do |f|
+ # f << "*Vertices #{Tag.all.size}\n"
+ # # tags = Tag.all.to_ary
+ # tag_index = {}
+ # Tag.all.each_with_index do |t,i|
+ # f << "#{i+1} \"#{t.name}\"\n"
+ # tag_index[t.id] = i+1
+ # end
+ #
+ # f << "*Edges #{edges_total}\n"
+ #
+ # hash.each do |id1, ids2Hash|
+ # ids2Hash.each do |id2, value|
+ # if id1 != id2
+ # f << "#{tag_index[id1]} #{tag_index[id2]} #{hash[id1][id2].to_f/(Math.sqrt(hash[id1][id1]*hash[id2][id2]))}\n"
+ # end
+ # end
+ # end
+ # end
+ #
+ # system("infomap --ftree #{graphPath} #{Rails.root.join(outDIR)}")
clusters = {childs: [], parent: nil}
tags = {}
@@ -84,8 +84,7 @@ namespace :tag do
it = clusters
ftree.each do |clusterId|
- # p it
- clusterId = clusterId.to_i
+ clusterId = clusterId.to_i - 1
if it[:childs][clusterId].nil?
it[:childs][clusterId] = {childs: [], parent: nil}
it[:childs][clusterId][:parent] = it
@@ -93,9 +92,127 @@ namespace :tag do
it = it[:childs][clusterId]
end
- it[:childs][leafId] = {id: tagId, rank: rank, name: name, parent: it}
- tags[tagId] = it
+ it[:childs][leafId-1] = {id: tagId, rank: rank, name: name, parent: it}
+ tags[tagId] = it[:childs][leafId-1]
end
end
+
+ def calculate_relevance(lo, close_tags)
+ rel = 0
+ lo.tags.each do |t|
+ close_tags.each do |cloT|
+ if cloT[:id] == t.id
+ rel += cloT[:rank]
+ end
+ end
+ end
+ return rel
+ end
+
+ def closest(tagId, tags)
+ clos = []
+ tags[tagId][:parent][:childs].each do |t1|
+ rank = (Math.log2(tags[tagId][:rank])-Math.log2(t1[:rank])).abs
+ if rank < 4
+ clos << { id: t1[:id], rank: rank}
+ end
+ end
+ normalize(clos)
+ end
+
+ def normalize(tags)
+ sum = 0
+ max = 0
+ tags.each do |t|
+ sum += t[:rank]
+ max = t[:rank] if t[:rank] > max
+ end
+ tags.each do |t|
+ t[:rank] = 1 - (t[:rank]/(max*1.05))
+ end
+ tags
+ end
+
+ def find_relevant_results(tagId, tags)
+ los = {}
+ close_tags = closest(tagId, tags)
+
+ p "==============="
+ p "close_tags"
+ close_tags.each {|ct| p Tag.find(ct[:id]).name+" | "+ct[:id].to_s+" | "+ct[:rank].to_s+" | "+tags[ct[:id]][:rank].to_s}
+ p "==============="
+
+ freq = cluster_frequency(close_tags)
+ LearningObject.all.each do |lo|
+ los[lo.id] = calculate_relevance(lo, close_tags)
+ # los[lo.id] = frequency_rank_global(lo, close_tags, freq)
+ # los[lo.id] = frequency_rank(lo, close_tags)
+ end
+
+ los = los.sort_by { |id, rel| rel }
+ lol = los.last(25).reverse
+ lol.each do |key, value|
+ puts "#{key}: #{value}"
+ end
+ lol
+ # lol.map {|lo| lo[0]}
+ end
+
+ def frequency_rank(lo, close_tags)
+ itf_sum = 0
+ wdf = 0
+ # t_size = lo.tags.size == 1 ? 2 : lo.tags.size
+ # wdf = 1/(Math.log2(t_size)) if lo.tags.size != 0
+ wdf = 1/(Math.log2(lo.tags.size)+1) if lo.tags.size != 0
+
+ lo.tags.each do |t|
+ close_tags.each do |cloT|
+ if cloT[:id] == t.id
+ itf_sum += cloT[:rank]*(Math.log2(Tag.all.size/t.taggings.size)+1)
+ end
+ end
+ end
+
+ return wdf*itf_sum
+ end
+
+ def cluster_frequency(cluster)
+ freq_cluster = 0
+ cluster.each do |t|
+ freq_cluster += Tag.find(t[:id]).taggings.size
+ end
+ freq_cluster
+ end
+
+ def frequency_rank_global(lo, close_tags, freq_cluster)
+ freq = 0
+ # rel = 0
+ lo.tags.each do |t|
+ close_tags.each do |cloT|
+ if cloT[:id] == t.id
+ freq += 1
+ # rel += cloT[:rank]
+ end
+ end
+ end
+
+ if lo.tags.size != 0
+ wdf = (Math.log2(freq+1)/(Math.log2(lo.tags.size)+1))
+ else
+ wdf = 0
+ end
+ itf = Math.log2(Tag.all.size/freq_cluster)+1
+ return wdf*itf#*rel
+ end
+
+ lol = find_relevant_results(22794, tags)
+
+ lol.each do |id, rank|
+ lo = LearningObject.find(id)
+ puts "-----"
+ p lo.id.to_s+": "+rank.to_s+" | "+lo.name
+ lo.tags.each {|t| print t.name+" - "+tags[t.id][:rank].to_s+" | "}
+ puts ""
+ end
end
end