Loading...
Please wait, while we are loading the content...
Similar Documents
Multidimensional Techniques for Privacy Preservation in Datasets
| Content Provider | Semantic Scholar |
|---|---|
| Author | Sushma, Nathani Kanaparthi, Priyanka |
| Abstract | Applications in commercial domains possess large datasets on individuals. This data includes private and sensitive information e.g. patient diseases, bank account details, organization structural details etc. When data mining techniques are applied on these applications the private and sensitive information of the subjects will be revealed. However, it is necessary to share the information in such a way that the identities of the individuals are not revealed. So it is necessary to anonymize the data. For this the quasi attribute set (attribute set that can be linked with original dataset to re-identify individuals) has to identified and anonymized. This paper presents the summary of various anonymization techniques Multidimensional generalization, Multidimensional suppression, Multidimensional Clustering and Multidimensional Cryptography to provide privacy for individuals. Any of these techniques can be applied to achieve privacy for individuals in a better manner. I. Introduction Knowledge discovery is the process of extracting valid, useful and understandable patterns from large databases or information repositories. Data mining plays a key role in knowledge discovery process. Many organizations publish micro data that includes public health and demographic information for different purposes. The private and sensitive details of users will be released when the data mining techniques are applied on these databases. The users require that this details to be confidential [1-3]. Although attributes that clearly identify individuals, such as Name, social security number are removed while releasing the data, these databases can sometimes be linked with other public databases like voter registration table etc on attributes such as Zip code, date of birth to re-identify individuals whose databases were supposed to remain anonymous. [4] This is shown in fig. 1 and fig. 2. Fig. 1, shows an employee Table with attributes employee number (empno), name, last name, date of birth (DOB) and salary for different employees. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.ijcst.com/vol24/3/nathani.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
