The　growing　popularity　of　online　social　networks　(OSNs)　in　recent　years　has　generated　a　lot　of　concern　on　personal　privacy.　One　approach　of　protecting　privacy　in　OSNs　is　to　intervene　in　the　flow　of　privacy　information,　making　the　study　of　the　dynamics　of　privacy　information　propagation　necessary　for　the　design　of　effective　privacy　protection　mechanisms.　Although　previous　work　on　information　propagation　has　produced　some　models,　these　models　are　not　adequate　for　privacy　information　since　they　do　not　reflect　the　main　characteristics　of　privacy　information.　In　this　paper,　we　propose　a　model　for　privacy　information　propagation.　We　first　analyze　the　structural　characteristics　of　privacy　information　and　then　design　the　model　by　incorporating　these　characteris-tics.　A　unique　feature　of　the　model　is　that　it　infers　the　privacy　attitudes　of　the　information　recipients　to　the　privacy　concerning　subject　implicated　in　the　privacy　information　to　determine　the　forwarding　decisions　of　the　recipients.　Thus,　by　mapping　the　heterogeneous　tendency　of　information　forwarding　by　the　recipients　to　a　limited　number　of　privacy　attitudes,　the　model　can　predict　the　decisions　on　forwarding　privacy　information　and　thus　describe　the　macroscopic　process　of　privacy　information　propagation.　Results　of　the　experiment　based　on　real　OSN　datasets　show　that　the　proposed　model　can　be　used　to　learn　both　the　scope　and　the　trend　of　privacy　information　propagation　in　OSNs,　demonstrating　the　importance　of　the　privacy　attitudes　of　recipients　on　privacy　information　propagation.　The　properties　of　the　model　are　also　studied　through　experiment　to　examine　the　impact　of　various　factors　on　privacy　information　propagation　in　OSNs.