US 12,224,985 B2
Edge data correctness and correctness spread determination
Mukundan Sundararajan, Bangalore (IN); Jignesh K Karia, Mumbai (IN); and Shilpa Shetty, Marathahalli (IN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Feb. 12, 2021, as Appl. No. 17/174,551.
Prior Publication US 2022/0263801 A1, Aug. 18, 2022
Int. Cl. H04L 9/40 (2022.01); G06N 20/00 (2019.01)
CPC H04L 63/0263 (2013.01) [G06N 20/00 (2019.01); H04L 63/1425 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method for providing edge data correctness and correctness spread determination, comprising:
identifying an incorrectness bias pattern in data received at an edge server of a network
through a filter with filter rules,
wherein the incorrectness bias includes spatial impurities in the data, and
wherein spatial impurities include anomalous data that does not follow a pattern of other data in the data received,
detecting, based on the incorrectness bias, anomalous data patterns, wherein the detected pattern indicates an identity spoofing issue;
simulating vulnerability data assessments for the spatial impurities;
identifying, based on the simulation, disguised data anomalies related to the spatial impurities from the incorrectness bias;
performing corrective measures to the data according to the disguised data anomalies that are identified;
learning and automatically updating the filter rules of the filter based on an extent of the incorrectness bias of the data and spread of the incorrectness bias of the data to reduce latency in future data validation of the data; and
creating security filter rules configured to mitigate incorrectness spread and incorrectness bias, wherein the security filter rules are varied across application and usage classes, and wherein a learning method provides a variance for the security filter rules.