227
Hội thảo Khoa học Quốc tế
...
PRIVACY PRESERVATION IN DISTRIBUTED ENVIRONMENT
USING NEGATIVE-ASSOCIATION RULE MINING ALGORITHM
Raghvendra Kumar
(Dept. of Computer Science and Engineering, LNCT Group of College,
Jabalpur, MP, India [email protected])
ABSTRACT
Data mining is method to extracting positive information from the huge quantity of database,
this contains dissimilar techniques to mine the database, mainly association rule mining, clustering,
classification etc. within this paper, tried with the concept of spatial negative association rule
mining (NARM), that extract the relationship among the different number of attributes in GIS
database to find the pattern between the attributes with high privacy and zero percentage of data
leakage. The most important purpose of this chapter is to extract the valuable pattern since the GIS
database; paper includes the following main steps
1. Consider the real geo-graphical database in distributed environments.
2. Perform the spatial negative-association rule mining (SNAR) in database.
3. Propose algorithm to find global relationship after SNAR mining.
4. Perform experiments with real geographic databases in order to validate the proposed
algorithm.
Keywords: Distributed Database, Privacy Preserving Technique (PPT), Negative Association
Rule Mining (NARM), Spatial Data Mining.
1. INTRODUCTION
Today is the era of internet, where the entire user wants to store and retrieve their private and
public information in the online storage survey [1] [2] [3]. When the data is stored in the server,
the problem is comes when the we consider to access their information, there are different number
of techniques are available in the field of data mining, like association rule mining, classification,
clustering etc. and also two main techniques first one is prediction, where the database admin
predicts the relationship between the end users or different number of attributes. And second one is
descriptive, where the database admin descript their useful information. In data mining techniques
association rule mining techniques that are very useful to find the relationship between the different
numbers of database. And second one is the clustering, where eliminate or grouping the attributes
according to their values, and last one is classification, where classify the attributes according to
their certain criteria, it may be age, education, etc.
1.1. Database: A Database [4] [5] [6] is the collection of data, where data represent useful
information gathered from the real world object. The system which manages these collected data is
called data base management system. This system is the necessity of the organization, enterprise etc.
Consider the example of an university database which has the information about faculty member
,staff member ,student ,course ,department etc. which will be change very frequently .There are
different type of database environment present in the network such as centralized and distributed.
In dissimilarity to the centralized data base model, the distributed data base model is fast but it
needs some extra effort for the privacy.
