SYSTEM AND METHOD BASED ON SLIDING-SCALE CLUSTER GROUPS FOR PRECISE LOOK-ALIKE MODELING

摘要

A graphical user interface showing relevant sliding scale cluster groups having a processor generating a ranking of Top N1 features of the seed dataset in order of frequency of their occurrence. The processor generates another ranking of Top N2 features of the seed dataset in order of frequency of their occurrence in correlation with their equivalent percentage in total dataset. The Top M co-related features is identified for each of the Top N (N1 + N2) features that are present for the seed dataset. The processor, for each one in M x N set, generates a sliding scale cluster via permutation of Top M features. The processor sorts each permutation based on closest occurrence match. The Seed Group Meta Bitmap Index is generated for the seed audience segment. For each cluster, the processor calculates the available amplification count from the Total Audience Bitmap Index, until the desired amplification is achieved.