US 12,443,884 B2
Device, system, and method for reducing machine learning bias in machine learning generated textual lists of time-stamped events
Chun Wen Ooi, Georgetown (MY); and Wooi Ping Teoh, Georgetown (MY)
Assigned to MOTOROLA SOLUTIONS, INC., Chicago, IL (US)
Filed by MOTOROLA SOLUTIONS, INC., Chicago, IL (US)
Filed on Dec. 29, 2022, as Appl. No. 18/090,702.
Prior Publication US 2024/0220850 A1, Jul. 4, 2024
Int. Cl. G06N 20/00 (2019.01)
CPC G06N 20/00 (2019.01) 20 Claims
OG exemplary drawing
 
1. A method comprising:
generating, via a computing device, via one or more machine learning algorithms, using sensor data, a textual list of time-stamped events associated with an incident, the sensor data related to the incident and generated by sensors, one or more of the time-stamped events in the textual list associated with respective machine learning confidence scores;
for a given time-stamped event, in the textual list, having a respective machine learning confidence score that is less than a threshold confidence score, redacting, via the computing device, the given time-stamped event in the textual list of time-stamped events and replacing the given time-stamped event in the textual list with a field for receiving input, the textual list of time-stamped events rendered at a display screen with the given time-stamped event in the textual list replaced with the field for receiving the input;
receiving, via the computing device, input at the field;
after receiving the input, un-redacting the given time-stamped event; and
rendering, via the computing device, at a display screen, the input received at the field and the given time-stamped event.