Multi-owner Encrypted Ranked Keyword Search Using Machine Learning Techniques
In the present era, the prevalence of network technology and cloud computing capabilities for multiple owner model has drawn much attention. Data owners outsource their data to the cloud and enjoy convenient services. However, extending single owner model
- PDF / 1,313,202 Bytes
- 14 Pages / 439.37 x 666.142 pts Page_size
- 40 Downloads / 237 Views
, Chungen Xu1(B)
, and Ali Zakir2
1 School of Science, Nanjing University of Science and Technology, Nanjing, China
{lailatulbadar,xuchung}@njust.edu.cn 2 School of Computer Science and Engineering, NJUST, Nanjing, China
[email protected]
Abstract. In the present era, the prevalence of network technology and cloud computing capabilities for multiple owner model has drawn much attention. Data owners outsource their data to the cloud and enjoy convenient services. However, extending single owner model into multiple owner model still have some issues, such as dimension reduction issue, high communication overhead, and efficient search in Searchable Symmetric Encryption (SSE) remain a challenging task. Integration of machine learning methods with the framework of searchable Encryption (SE) is a diverse way to solve these problems. In this paper, we developed Multi-owner encrypted ranked keyword search using Machine learning techniques(MERS-ML). Our Scheme utilized principal component Analysis (PCA) model to reduce high-dimensional data into low-dimensional codes and enabled low-overhead system maintenance. K-means clustering approach is used to solve the problem of the quality of different document of multiple owner model. To achieve fast query and efficient search relatively close to O(log N), we designed search balanced index tree. Besides, attribute-based encryption is used to achieve convenient key management as well as authorized access control. Compared with previous work, our scheme provide adequate privacy protection, improved search efficiency, and introduce low overhead on computation and storage. Keywords: Searchable encryption · Ranked keyword search · Multi owner model · Binary index tree · Principle component analysis
1 Introduction Background. As a promising computing paragon, cloud computing has become a hot topic for research and industrial communities. It brings huge benefits to data owner such as, provide inclusiveness, flexibility, scalability, and rapid retrieval of data. Due to its highly desirable features, organizations, as well as individuals, are influenced to outsource their data onto the cloud server to accomplish ease and low management cost. Unfortunately, cloud computing is facing many problems and challenges during the transaction and storage of the outsourced data. Outsourced data onto the cloud contain sensitive information such as medical records, organization’s financial records etc. Illegitimate use of client personal information or reveal any sort of private data is © Springer Nature Singapore Pte Ltd. 2020 S. Yu et al. (Eds.): SPDE 2020, CCIS 1268, pp. 399–412, 2020. https://doi.org/10.1007/978-981-15-9129-7_28
400
L. T. Badar et al.
seemingly occur as data is stored in third-party cloud server. For the safety, security, and privacy of data, researchers have been paid much attention to the growing security incidents in cloud computing. To solve these issues, data must be encrypted before outsourced into the cloud server. Related work and Challenges. Song et al. [1] p
Data Loading...
