In today’s world, various methods are being deployed to establish a secure communication and fast keyword searching over a network. The existing semantic secure Public-key Encryption with Keyword Search (PEKS) takes linear search time with the total number of ciphertexts which makes it time consuming. Hence retrieval of keywords from the large-scale database is restricted. To tackle with this drawback, our paper proposes complete system that implements the fast keyword search using Public-key Ciphertext with Hidden Structures (FKSCH) for fast keyword search without sacrificing the linguistic security of encrypted keywords. The searchable ciphertext keywords are structured by hidden relations, which reveals only a part of information to match the ciphertext efficiently preserving the privacy of the data. Hence the time taken by FKSCH for the keyword search depends on actual number of ciphertext containing queried keyword rather than the whole ciphertext in the cloud database. Our scheme satisfies the standard rules of Random Oracle model and proves it to be better than standard model.