Abstract: Many data owners outsource their data management to cloud servers for simplicity and lower expenses. Encrypting critical data before outsourcing ensures its privacy. Law requires this. Thus, many data-use tactics, such as keyword-based document retrieval, are antiquated. Our method decrypts cloud-stored data without putting users at risk. This method allows simultaneous adjustments. The system allows document deletion and addition. Vector space models and TF-IDF models help indexing and query creation. Using "Greedy Depth First Search," you may search for terms in our tree-based database. Users may search for multiple keywords in the order they appear in results. Secure kNN encrypts both index and query vectors. Encrypting the index and query vectors allows for accurate relevance scoring. This method encrypts index and query vectors. Phantom words conceal search results and protect the index vector against statistical attacks. This safeguards the index. Our tree-based index structure speeds up searches using the suggested strategy. Quickly removing and inserting files are other benefits. The index's tree-like structure enables this. The proposed approach will be tested thoroughly.

Keywords: Multi Keyword Search, Encrypted, Cloud, Data, Cloud Servers.

| DOI: 10.17148/IARJSET.2022.96121