The purpose of the paper is to address the issue of mining fuzzy association rules in databases designed using Entity-Relationship (ER) Models at multiple levels to find the patterns of the people involved in cybercrime with respect to their age and employment to the degree of crime.. The study aims to extend the previous developed algorithms Extended Apriori and Apriori star to discover a new algorithm. The study will help in standardizing algorithms for finding appropriate results from database tables containing fuzzy data.
The proposed study focuses on developing an algorithm that will use fuzzy logic for finding fuzzy association rules from ER models for Cyber Crime. The study will help to find the patterns and identify the category of people involved in cybercrime. Analysis of the transaction data which includes the sample personal data and the regions where they live converted into fuzzy taxonomic structures reflecting age and region involved in cybercrime. The study will focus on finding rules such as: [Employed (medium), Age = Young] => [Violent Crimes (low)] which implies that test sample of the age group 20-30 and 30-40 who is employed might engage in cyber crime with qualitative relevance low. In the rule mentioned the young, medium and low belong to three different entities designed using ER models. Another example of such a rule can be Height = Tall Game = Basket Ball which implies that a person with the height 5’ 7’’ – 5’ 9’’ and 6’ 0 “+ might opt for Basket Ball game where group 5’ 7” – 5’ 9” partially belongs to Tall with degree Tall5’7”–5’ 9”and both Height and Game belong to two different entities. In this example the attributes Height of the candidate table will be first converted into fuzzy taxonomic structures respectively reflecting partial belonging of one item to another.
The paper discovers a new algorithm Extended Apriori Star to find such kind of rules. The fuzzy extensions that will be presented in this study will enable us to discover not only crisp generalized association rules but also fuzzy generalized association rules when databases consisting of several tables organized in a schema within the framework of fuzzy taxonomic structures and helps to show the patterns of the people involved in Cyber Crime. Strong Association rules between items of fuzzy nature existing in multiple tables can be calculated that will undoubtedly help in understanding things in broad spectrum.
Reference:
Praveen Arora, Ram Kumar, Ashwani Kush, “Mining fuzzy generalized association rules for ER models”.International Journal of Information Technology and Knowledge Management (IJITKM), 2008, vol-1, issue-2, pp 191-198.
Image - (via rodic)
0 comments:
Post a Comment