Securing information system filtering from malicious attacks has become an important issue with increasing popularity of information system filtering. Data mining is the analysis of observational data sets to find unsuspected relationships and to summarize the data novel ways that are both understandable and useful to data owners. Information systems are entirely based on the input provided by users or customers, they tend to become highly prone to attacks. To prevent such attacks several mechanisms can be used. In this paper, we show that the unsupervised clustering one of the data mining technique can be used for attack detection for all types of attacks. This method is based on computing detection attributes modeled on basic descriptive statistics. Our study showed that attribute based unsupervised clustering algorithm can detect spam users with high degree of accuracy and fewer misclassified genuine users regardless of attack strategies.

Data mining clustering information system filtering