CPC G06F 16/215 (2019.01) [G06F 16/2379 (2019.01); G06F 21/6245 (2013.01); G06Q 10/10 (2013.01); H04L 63/02 (2013.01)] | 18 Claims |
1. A computer implemented method for updating customer data that includes a plurality of electronically stored contact records that each include contact information for a respective individual contact, comprising:
applying automated filtering to the customer data to identify one or more duplicate contacts that fall within defined filtering criteria, wherein the defined filtering criteria include similarities in separate individual contact data found in multiple contact records, and the similarities are identified through logic based comparison algorithms;
generating a multi-dimensional feature vector representing each identified duplicate contact;
performing cluster analysis on the resulting feature vectors to reduce the identified duplicate contacts to a smaller candidate list of one or more identified duplicate contacts;
providing information about the duplicate contacts included in the smaller candidate list to a decision making authority for additional filtering; and
receiving feedback derived from the decision making authority in respect of the provided information, and based on the feedback, automatically updating the customer data in respect of the duplicate contacts included in the smaller candidate list.
|