US 11,741,384 B2
Adaptable systems and methods for discovering intent from enterprise data
Tilak B Kasturi, Palo Alto, CA (US); Hieu Ho, San Jose, CA (US); and Aniket Dalal, Santa Clara, CA (US)
Assigned to PREDII, INC., Palo Alto, CA (US)
Filed by Predii Inc., Palo Alto, CA (US)
Filed on Oct. 14, 2019, as Appl. No. 16/601,525.
Claims priority of provisional application 62/745,285, filed on Oct. 13, 2018.
Prior Publication US 2020/0118014 A1, Apr. 16, 2020
Int. Cl. G06N 20/00 (2019.01); G06N 5/045 (2023.01); G06F 16/28 (2019.01); G06F 40/30 (2020.01); G06N 7/01 (2023.01)
CPC G06N 5/045 (2013.01) [G06F 16/283 (2019.01); G06F 40/30 (2020.01); G06N 7/01 (2023.01); G06N 20/00 (2019.01)] 15 Claims
OG exemplary drawing
 
1. A method for analyzing data, comprising:
defining an intent language model for domain specific meaning behind historical enterprise data produced during operation of an enterprise, wherein the historical enterprise data includes data generated by equipment during operation, issues or symptoms reported about the equipment, findings and observations reported by one or more human experts in one or more service records including recommended actions, repairs, parts or recommendations, and wherein the historical enterprise data further includes data associated with a problem;
applying the historical enterprise data to build the intent language model;
extracting intent element features of interest from the historical enterprise data to generate domain specific intent metadata, wherein the domain specific intent metadata includes a root cause of the problem, wherein the root cause is an intent extracted from the historical enterprise data based on the problem; and
storing the domain specific intent metadata into a database.