Correlations and Associations Analysis for Identification and Prediction of Relationships and Their Activeness

This paper presents a novel approach to build a model for finding relationships, building associations, and for the prediction of activeness. The model helps in analyzing how cyber space has transformed human behavior and relationships. The model uses clustering algorithms and fuzzy logic to determine the relationship patterns of number of selected users in a network environment. The behavior of each of these users is determined by feature selection. Several features are used to predict the activeness of their relationship. Time series based model dynamically updates the prediction rules based on the current happenings. Feature relationships are studied under feature cluster analysis, which is an attempt to define clusters by measuring the interaction between the nodes in the environment then updating the model based on this information. Features analysis involves the analysis of different attributes in an email. As these features are increased the more relationships and behavior patterns can be analyzed and build