Algorithm 1: Creation of tag hierarchy Result: Popular tags structured in a hierarchy create root ordered tags←root for each tag t do t if t popularity popularity threshold then t。←mnar( tatar,vtag∈ ordered tags) if ol overlap threshold th if rightcontert? (to ancestors, t) then I to children←t root, children←toc end else if Itol<t then f rightcontert? (to children, t) then t parent to parent children←to create a copy toc of to root, children end else end end else root. children←t end end 2.3 Applying Knowledge on Keywords from Wordnet ordnet 8(wordnet. princeton. edu) is a large lexical database for the Er glish language developed and maintained by Cognitive Science Laboratory at Princeton University. It groups nouns, verbs, adjectives and adverbs into sets of cognitive synonyms(synsets). Moreover, it provides conceptual-semantic andAlgorithm 1: Creation of tag hierarchy Data: a folksonomy Result: Popular tags structured in a hierarchy create root ordered tags ← root for each tag t do t.popularity = |t.users| |overallusers| if t.popularity < popularity threshold then continue end to ← max( t ∩ tagx, ∀tagx ∈ ordered tags ) if |to| |t| > overlap threshold then if |to|  |t| then if rightcontext?(to.ancestors, t) then to.children ← t end else create a copy toc of to root.children ← toc to.children ← t end end else if |to|  |t| then if rightcontext?(to.children, t) then t.parent = to.parent t.children ← to end else create a copy toc of to t.children ← toc root.children ← t end end else to.sibling ← t end end else root.children ← t end end 2.3 Applying Knowledge on Keywords from Wordnet Wordnet [8] (wordnet.princeton.edu) is a large lexical database for the En￾glish language developed and maintained by Cognitive Science Laboratory at Princeton University. It groups nouns, verbs, adjectives and adverbs into sets of cognitive synonyms (synsets). Moreover, it provides conceptual-semantic and