US 12,235,855 B2
Data discovery solution for data curation
Andy Tae Ahn, Hillsborough, CA (US)
Assigned to PRAXI DATA, INC., Hillsborough, CA (US)
Filed by PRAXI DATA, INC., Hillsborough, CA (US)
Filed on Jan. 26, 2023, as Appl. No. 18/160,254.
Application 18/160,254 is a continuation of application No. 17/062,242, filed on Oct. 2, 2020, granted, now 11,567,949.
Application 17/062,242 is a continuation of application No. 16/514,983, filed on Jul. 17, 2019, granted, now 10,795,899, issued on Oct. 6, 2020.
Claims priority of provisional application 62/699,518, filed on Jul. 17, 2018.
Prior Publication US 2023/0273928 A1, Aug. 31, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/2457 (2019.01); A63F 13/79 (2014.01); G06F 9/50 (2006.01); G06N 20/00 (2019.01)
CPC G06F 16/24578 (2019.01) [A63F 13/79 (2014.09); G06F 9/5005 (2013.01); G06F 16/24575 (2019.01); G06N 20/00 (2019.01)] 24 Claims
OG exemplary drawing
 
1. A method of generating a set of candidate patterns that train a model comprising:
receiving training input, including:
transmitting, via a matching application, a first matching question to a first player of a plurality of players, wherein the first matching question is configured to solicit a response that matches a first data element of a dataset to a first tag;
receiving, via the matching application, a first player response that matches the first data element to the first; and
storing a first pattern to a database comprising a corpus of patterns, the first pattern corresponding to the first player response;
validating the first pattern within the corpus of patterns, including:
transmitting, via the matching application, a second matching question to the first player, wherein the second matching question comprises a validator question corresponding to at least one past answer;
receiving, via the matching application, a second player response to the second matching question; and
generating a reliability rating for the first player based on consistency of the second player response to the second matching question with the at least one past answer;
weighting the first pattern in the corpus of patterns based on the reliability rating of the first player; and
training an artificial intelligence model using the first pattern weighted based on the reliability rating of the first player.