Analyzing　similarity　of　access　control　policies　is　very　important　in　many　application　sceneries,　for　example,　conducting　similarity　analysis　to　reduce　policy　scale　in　cloud　services　composition,　or　through　similar　policies　to　find　cloud　services　which　have　same　function.　Existing　methods　of　policy　similarity　analysis　are　mainly　based　on　logical　reasoning　or　Boolean　function　comparison,　which　are　computationally　expensive　and　do　not　scale　well　for　large　heterogeneous　distributed　cloud　environments.　In　this　paper,　we　propose　a　lightweight　approach　for　analyzing　the　similarity　of　heterogeneous　policy　in　cloud　environment,　called　Saphena.　In　Saphena,　paired　attributes　are　used　to　unify　and　simplify　heterogeneous　policies　and　the　similarity　of　attribute　is　calculated　by　utilizing　semantic-based　analysis　method.　The　similarity　of　policies　is　obtained　by　digging　the　similarities　of　all　attributes　in　policies　and　controlling　the　corresponding　weighting　factor.　We　design　a　threshold　that　is　used　to　judge　whether　similarity　among　policies.　Cloud　service　providers　can　set　threshold　through　interface　flexibly　based　on　their　difference　request.　Comprehensive　experiments　have　been　conducted,　which　demonstrate　the　effectiveness　of　the　proposed　method　by　the　comparison　with　the　classic　cosine　similarity　analysis　arithmetic.