US 11,940,867 B2
Method for managing a plurality of events
Prasad Vyavahare, San Joe, CA (US); Swati Choksi, San Joe, CA (US); Silvia Veronese, San Jose, CA (US); Roger Brooks, San Jose, CA (US); and Zainab Jamal, San Jose, CA (US)
Assigned to GUAVUS Inc., San Jose, CA (US)
Filed by GUAVUS INC., San Jose, CA (US)
Filed on Feb. 2, 2023, as Appl. No. 18/104,884.
Application 18/104,884 is a continuation in part of application No. 16/942,038, filed on Jul. 29, 2020, granted, now 11,641,304.
Prior Publication US 2023/0185650 A1, Jun. 15, 2023
Int. Cl. G06F 11/07 (2006.01); G06F 11/30 (2006.01); G06F 11/32 (2006.01)
CPC G06F 11/0781 (2013.01) [G06F 11/3006 (2013.01); G06F 11/327 (2013.01)] 14 Claims
OG exemplary drawing
 
1. A method for managing a plurality of events, wherein each event comprises physical attributes and logical attributes, the method comprising the steps of:
creating tuples, wherein each tuple is an identifier for a set of logical attributes to events having all the same logical attributes;
providing a set of hierarchized relations between tuples, wherein combinations of two tuples form a pair of tuples, by means of an unsupervised machine learning pipeline algorithm, wherein parent-child relations are provided between tuples, by:
creating a plurality of binarized co-occurrence matrices, each co-occurrence matrix reflecting different time intervals, wherein each column corresponds with a tuple and each row corresponds with a time window, so each matrix entry at a tuple column and a time window row represents that at least one event corresponding to the tuple associated with the tuple column appears in each time window associated with the time window row;
successively applying a heuristic function to each matrix entry of said plurality of co-occurrence matrices to obtain a co-occurrence probabilistic score for each pair of tuples, wherein the probabilistic score indicates the probability that one tuple of the pair, referred to as child tuple, co-occurs with the other tuple of the pair, referred to as parent tuple; and
using the probabilistic score of each pair of tuples to quantify the strength of the parent-child relations;
classifying the tuples in families, each family contains all the tuples related according to the parent-child relations provided by the unsupervised machine learning pipeline algorithm;
identifying the parent tuple of each family, defined as a tuple that has at least one child and has no parent;
extracting instance tuples associated with each tuple in each tuple family thereby creating instance families;
presenting the parent tuple instances of each instance family, together with the physical attributes of the events associated to each parent instance tuple each instance family.