Currently, searchable encryption has attracted considerable attention in the field of cloud computing. The existing research mainly focuses on keyword-based search schemes, most of which support the exact matching of keywords. However, keyword-based search schemes ignore spelling errors and semantic expansions of keywords. The significant drawback makes the existing techniques unsuitable in cloud computing as it greatly affects system usability and can not completely satisfy the users' search intentions. In this paper, we propose an effective fuzzy semantic searchable encryption scheme (FSSE) that supports multi-keyword search over encrypted data in cloud computing. In our scheme, we exploit a keyword fingerprint generation algorithm to generate a fingerprint set of the keyword dictionary and a fingerprint of the query keywords, and employ Hamming distance to quantify keywords similarity. Based on the proposed fingerprint generation algorithm and Hamming distance, we realize fuzzy search. Furthermore, we utilize the semantic expansion technique to expand query keywords and calculate the semantic similarity between the query keywords and the expanded word of the query keywords to achieve the semantic search. To improve the search efficiency, we construct an inverted index structure and use the vector intersection matching as well as short-circuit matching operations to effectively filter irrelevant documents. The theoretical analysis and experimental results demonstrate that our proposed scheme satisfies the security guarantee of searchable encryption, enhances system usability, and is more efficient in comparison with the state of the art schemes. INDEX TERMS Searchable encryption, cloud computing, fuzzy semantic search, multi-keyword search. I. INTRODUCTION

Cyber physical system (CPS) is facing enormous security challenges because of open and interconnected network and the interaction between cyber components and physical components, the development of cyber physical systems is constrained by security and privacy threats. A feasible solution is to combine the fully homomorphic encryption (FHE) technique to realize the efficient operation of ciphertext without decryption. However, most current homomorphic encryption algorithms only support limited data types, making it difficult to be widely applied in actual environment. To address this limitation, we propose a parallel fully homomorphic encryption algorithm that supports floating-point numbers. The proposed algorithm not only expands the data types supported by the existing fully homomorphic encryption algorithms, but also utilizes the characteristics of multi-nodes in cloud environment to conduct parallel encryption through simultaneous group-wise ciphertext computations. The experimental results show that, in a 16-core 4-node cluster with MapReduce environment, the proposed encryption algorithm achieves the maximum speed-up exceeding 5, which not only solves the limited application problem of the existing fully homomorphic encryption algorithm, but also meets the requirements for the efficient homomorphic encryption of floating-point numbers in cloud computing environment.

With the widespread application of big data, privacy-preserving data analysis has become a topic of increasing significance. The current research studies mainly focus on privacy-preserving classification and regression. However, principal component analysis (PCA) is also an effective data analysis method which can be used to reduce the data dimensionality, commonly used in data processing, machine learning, and data mining. In order to implement approximate PCA while preserving data privacy, we apply the Laplace mechanism to propose two differential privacy principal component analysis algorithms: Laplace input perturbation (LIP) and Laplace output perturbation (LOP). We evaluate the performance of LIP and LOP in terms of noise magnitude and approximation error theoretically and experimentally. In addition, we explore the variation of performance of the two algorithms with different parameters such as number of samples, target dimension, and privacy parameter. Theoretical and experimental results show that algorithm LIP adds less noise and has lower approximation error than LOP. To verify the effectiveness of algorithm LIP, we compare our LIP with other algorithms. The experimental results show that algorithm LIP can provide strong privacy guarantee and good data utility.

No abstract