With　the　deep　integration　of　the　Internet　of　Things　(IoT)　and　cloud　computing,　cloud-oriented　IoT　is　embraced　as　an　important　paradigm　for　efficiency　and　productivity.　On　the　other　hand,　it　is　also　becoming　an　increasingly　attractive　target　for　cybercriminals,　who　attempt　to　breach　data　security　and　privacy.　As　a　potential　and　promising　solution　to　secure　data,　ciphertext-policy　attribute-based　keyword　search　(CP-ABKS)　can　provide　both　fine-grained　keyword　search　and　access　control　over　the　encrypted　data.　However,　prior　CP-ABKS　schemes　either　fail　to　support　lightweight　computation　or　lack　of　policy　protection.　In　this　paper,　with　offline　computation　and　inner　product　encryption,　we　propose　a　lightweight　CP-ABKS　scheme　with　policy　protection,　such　that　the　encrypted　data　can　be　efficiently　retrieved　and　accessed　by　data　users　in　a　fine-grained　manner　without　leaking　any　sensitive　information.　We　prove　the　correctness　of　the　proposed　scheme　and　its　security　in　the　standard　model　under　the　Decisional　Bilinear　Diffie-Hellman　(DBDH)　assumption.　We　also　implement　our　proposed　scheme　to　demonstrate　its　efficiency.