Abstract: An increasing popularity of cloud computing, an ever increasing number of information owners are inspired to
outsource their information to cloud servers for awesome comfort and reduced cost in information administration. However,
sensitive data should be encrypted before outsourcing for privacy requirements, which obsoletes data utilization like keywordbased
document retrieval. In this paper, we introduce a secure multi-keyword ranked search scheme over encoded cloud
information, which at the same time supports dynamic update operations like cancellation and addition of archives. In
particular, the vector space demonstrate and the generally utilized TF-IDF show are joined in the record development and
inquiry age. In this paper build a unique tree-based file structure and propose an ”Greedy Depth-first Search” algorithm to
give productive multi-keyword ranked search. The protected kNN algorithm is used to encode the file and inquiry vectors,
what’s more, in the interim guarantee precise importance score count between encrypted index and query vectors. Keeping in
mind the end goal to oppose attacks, ghost terms are added to the index vector for blinding indexed lists. Because of the
utilization of our unique tree-based index structure, the proposed plan can accomplish sub-direct inquiry time and manage the
cancellation and addition of records adaptable. Extensive experiments are led to show the effectiveness of the proposed scheme.
KEYWORDS: TF-IDF, multi-keyword, KNN