Authenticated CRF Based Improved Ranked Multi-Keyword Search for Multi-Owner Model in Cloud Computing

Abstract

As advances in cloud computing enhanced, more data owners are getting attracted towards it for storing their data over cloud storage to reduce the data management and performance overhead. It is a flexible and economical solution for people. But in multi data owner model it arises various issues such as maintaining privacy of data owner, improving the security of data, reducing complexity of search over encrypted data etc. It is a challenging problem to allow the multi keyword search along with most relevant search results in multi data user and huge document model in cloud. To provide the solution to above mentioned problems, this paper designs the privacy preserving ranked multi keyword search system for multiple data owners (PRMSM). This system allows the secure search with trap door by using secure search protocol and improves the search efficiency by using Conditional Random Field (CRF) method. This method is a combination of tokenization, POS tagging and score calculation. It is used to remove the unwanted data along with reducing the index size which leads to improved search efficiency. The experimental results proves that the system maintains the security, privacy, authenticity along with improved search efficiency and ranking of results to improve the experience of searching. This system also supports the feature of user revocation.