29-08-2014, 04:56 PM
DESIGN AND IMPLEMENTATION OF DATA SANITAZATION TECHNIQUE FOR EFFECTIVE FILTERING WITH ENCHANCED MEDICAL SUPPORT SYSTEM IN CLOUD
Design and Implementation.doc (Size: 663.5 KB / Downloads: 8)
ABSTRACT
In the EXISTING SYSTEM, thousands of textual documents are publicly published every day. Even though methods to assist the sanitization process have been proposed, most of them are focused on the detection of specific types of sensitive entities for concrete domains, lacking generality and requiring user supervision. In the PROPOSED SYSTEM, We are developing this Project for Medical Purpose. Here we use the Cloud Server as a main Server, where all the Data from the Users are Stored. We design this system using Registered Doctors, Paid and unpaid users. Data Sanitization is achieved by Three Process. 1. Entity Generalization-Preserving the Privacy data with its semantics. 2. Entity Swapping is used to Reduce the Document Size. 3. Noise Addition: an entity substituted by another similar one extracted from another repository. In the MODIFICATION Process, Paid users are only allowed to access the Doctor’s Opinion/Suggestion/ Prescriptions. Registered Doctors can only Reply to the User’s/ Patients.
INTRODUCTION
In The context of the Information Society, thousands of documents potentially containing sensitive information are made public or available for third parties daily for a variety of reasons. Governments that publish documents in response to Freedom of Information requests or medical data like electronic health care records, which are made available due to their usefulness for clinical research are examples of this situation. Moreover, in recent years, the emergence of the Cloud has represented a fundamental change in the way information technology services are designed and deployed in business and governments. In fact, the use of cloud environments has predominantly focused on information sharing and communications. More specifically, the use of document-sharing applications is one of the main opportunities for the cloud computing
EXISTING SYSTEM
In the EXISTING SYSTEM, there is no big implementation regarding Security was introduced in Cloud Computing. Also there are thousands of textual documents are publicly published every day. Even though methods to assist the sanitization process have been proposed, most of them are focused on the detection of specific types of sensitive entities for concrete domains, lacking generality and requiring user supervision
PROPOSED SYSTEM
To overcome this drawback, We are developing this Project for Medical Purpose. Here we use the Cloud Server as a main Server, where all the Data from the Users are Stored. We design this system using Registered Doctors, Paid and unpaid users. Data Sanitization is achieved by Three Process. 1. Entity Generalization-Preserving the Privacy data with its semantics. 2. Entity Swapping is used to reduce the Document Size. 3. Noise Addition: an entity substituted by another similar one extracted from another repository.
CONCLUSION
In this paper, an automatic text sanitization method has been proposed. It relies on the theoretical foundations of the information theory and a corpus as global as the Web to offer a general-purpose solution that can be applied to heterogeneous textual data (and not only NEs ). Contrary to methods based on k-anonymity models, which deal with groups of documents with a similar structure/topic in order to swap and replace sensible entities, our method is able to sanitize each document independently. Moreover, it offers a flexible and intuitive way (in comparison with abstract numerical parameters) to configure the sanitization degree, based on domain- specific linguistic features. Finally, special care has been put in the preservation of document’s utility, as a function of its semantics. General-purpose knowledge sources have been used to reduce the amount of information given by document terms while maintaining, up to a degree, their semantics. Evaluation results, obtained for entities of different domains, sustained the theoretical premises, showing a high detection recall in comparison with general-purpose approaches based on trained classifiers. Document’s utility was also better retained, in comparison with methods based on term suppression, with values close to the ideal sanitization and coherent with sanitization thresholds.