Multi-Keyword Search Based on Top-k Cosine Similarity with Co-Ordinate Matching

摘要

Cloud Computing provides enormous computing power as well as scalable storage capacities to support a huge amount of data owners for outsourcing their data on a cloud server. However datasets contain sensitive information, for privacy concern documents are encrypted before outsourcing. Large numbers of data users give Multi-keyword queries to a cloud server and retrieve relevant documents. In this paper, unauthorized access is mitigated by encrypting the data using Hybrid Rivest Shamir Adleman (RSA) and Advanced Encryption Standard (AES) algorithm before outsourcing. Multi-keyword search is performed based on the Top-k Cosine Similarity Algorithm with Co-ordinate Matching. For improving relevance between queries and documents, Association Analysis is done based on the idea of Top-k cosine similarity algorithm. Then Coordinate matching is used to capture the top-k document for the query. The experimental results show that our proposed system achieves 60% improvement in retrieving top-k documents compared to Cipher Text Model (BMTS) and 20% compared to Ranked Searchable Symmetric Encryption (RSSE). It also provides better performance compared to the existing system in terms of encryption time and accuracy.