With　the　increasing　development　of　business　process　techniques　in　many　enterprises　and　organizations,　business　process　recommendation　has　become　a　dramatic　research　area　of　business　process　management　techniques.　However,　there　are　some　problems　in　current　business　process　recommendation　approaches,　which　rely　on　simple　metrics　such　as　structural,　textual,　or　behavioral　similarity,　or　which　generally　rely　on　specific　structures　without　typical　metadata　features.　To　improve　the　availability　of　process　description　and　recommendation,　an　approach　for　recommending　processes　is　proposed　based　on　metadata.　The　concept　of　business　process　description　framework　(BPDF)　is　first　constructed　based　on　MFI-5.　According　to　BPDF,　similarity　feature　set　(SFS)　of　the　process　is　defined.　The　processes　are　further　identified　and　quantified　using　SFS,　so　that　the　process　vectors　are　obtained.　And　the　similarity　measure　algorithm　is　utilized　to　calculate　the　similarity　between　any　two　vectors,　and　the　similarity　matrix　of　the　processes　can　then　be　extracted.　According　to　the　results　of　processes　similarity　measurement,　these　processes　are　ranked　to　provide　support　for　process　recommendation.　An　empirical　experiment　shows　that　the　proposed　approach　can　be　effectively　applied　to　actual　scenarios　of　process　recommendation.