The major use of the distributed computing is when the users can store their data in any form on remote server cloud while these data accessibility is also permitted for the authorized users. Since the secured data and the personal information are inclined to changing of data, these are highly protected in the cloud servers. This is done by the process of encoding of the data. The data must be secured in terms of how they could be transformed easily if they aren’t encrypted. This project aims at addressing this issue by working on the fine grained multi word data sets in the information that is encoded. This is hence achieved by a three-fold process. Initially, we present the importance scores and inclination elements upon watchwords which empower the exact catchphrase look and customized client experience. Second, we add to a viable and exceptionally productive multi-watchword look plan. The proposed plan can bolster convoluted rationale seek the blended "AND", "OR" and "NO" operations of watchwords. Third, we advance utilize the arranged subword references method to accomplish better productivity on file building, trapdoor creating and question. Ultimately, we dissect the security of the proposed plans as far as classification of records, security insurance of file and trapdoor, and unlink capacity of trapdoor. The security investigation and the results that prove to be explanatory show that the plans proposed can accomplish a level of security which is different from the present levels and also provides with a useful and productive output as required and expected.