What is it about?

In this work, we present a superior damage assessment and recovery algorithm that is centered on agents. Our hybrid lightweight approach is based on clustering database transactions based on a given set of criteria using graphs to keep track of transactional dependencies.

Featured Image

Why is it important?

Our model allows for: (1) parallel information processing—which makes recovery more effectual, (2) separation of concerns—which makes it easier to maintain a given data structure, (3) attack/problem isolation where a malicious transaction will be isolated from the remaining unharmed parts of database (the undamaged parts of the database can remain ‘live’ and there is no need to take them offline, and (4) easier scaling as bottlenecks are diminished.


Writing this article was a great pleasure as it has co-authors with whom I have had long standing collaborations.

Dr. Ramzi A. Haraty
Lebanese American University

Read the Original

This page is a summary of: Information reconciliation through an agent-controlled graph model, Soft Computing, March 2020, Springer Science + Business Media,
DOI: 10.1007/s00500-020-04779-x.
You can read the full text:




The following have contributed to this page