Data Mining and Knowledge Engineering

M. P. Sruthi ¹, S. Sinthiya ¹, B. Gomathy ¹

Affiliations
1 Department of Computer Science and Engineering, Bannari Amman Institute of Technology, Sathyamangalam, Erode, IN

Abstract

Personal data leads to a trend that data owners adopt to remotely outsource their data to clouds for the enjoyment of the high-quality revival and storage service without worrying the load of local data management and maintenance. However, protected share and search for the outsourced data is a challenging task, which may easily incur the leakage of sensible personal information. Efficient data sharing and searching with security is of essential importance. This paper, for the first time, introduces a searchable attribute-based proxy re-encryption system. When we compare to existing systems only aiding both searchable attribute-based functionality and attribute-based proxy re-encryption, our new primitive supports both abilities and gives flexible keyword update service. Particularly, the system enables a data owner to efficiently share his data to a named group of users matching a sharing policy and the data will maintain its searchable property but also the respective search keyword can be uploaded after the data sharing. The new mechanism is suitable to many real-world applications, such as electronic health record systems. It is proved that chosen cipher text is secure in the random oracle model. To achieve fine-grained and efficient scalable data access control for PHRs, we grease attribute based encryption (ABE) techniques to encode each patient's PHR file. The highest degree of patient privacy is assured simultaneously by using multi-authority ABE. Our scheme enables dynamic modification of access policies and file attributes, guide adequate on-demand user/attribute revocation and break-glass access based on emergency scenarios. Extensive detailed and empirical results are presented which shows the security, scalability and efficiency of proposed scheme.

Keywords

Patient, Records, Sharing, Attribute Based Encryption.

Full Text

     

M. Priyanka ¹, N. Subha ¹, B. Gomathy ¹

Affiliations
1 Department of Computer Science and Engineering, Bannari Amman Institute of Technology, Sathyamangalam, Erode, IN

Abstract

Data mining is the practice of extracting the useful information from the frequently occurring patterns in the observational datasets available. In data mining, classification is the process of bringing out the models called classifiers to predict the class labels. This paper gives a comprehensive review of the classification techniques in data mining such as Support Vector Machine (SVM), Decision tree and Linear Regression.

Keywords

Classification, Datasets, Decision Tree, Linear Regression, SVM.

Full Text