US 11,972,220 B2
Enhanced logits for natural language processing
Ying Xu, Albion (AU); Poorya Zaremoodi, Melbourne (AU); Thanh Tien Vu, Herston (AU); Cong Duy Vu Hoang, Wantirna South (AU); Vladislav Blinov, Melbourne (AU); Yu-Heng Hong, Carlton (AU); Yakupitiyage Don Thanuja Samodhye Dharmasiri, Melbourne (AU); Vishal Vishnoi, Redwood City, CA (US); Elias Luqman Jalaluddin, Seattle, WA (US); Manish Parekh, San Jose, CA (US); Thanh Long Duong, Melbourne (AU); and Mark Edward Johnson, Castle Cove (AU)
Assigned to Oracle International Corporation, Redwood Shores, CA (US)
Filed by Oracle International Corporation, Redwood Shores, CA (US)
Filed on Nov. 29, 2021, as Appl. No. 17/456,687.
Claims priority of provisional application 63/119,449, filed on Nov. 30, 2020.
Prior Publication US 2022/0171946 A1, Jun. 2, 2022
Int. Cl. G06F 40/35 (2020.01); G06F 40/205 (2020.01); G06F 40/253 (2020.01); G06N 3/08 (2023.01); H04L 51/02 (2022.01)
CPC G06F 40/35 (2020.01) [G06N 3/08 (2013.01); H04L 51/02 (2013.01); G06F 40/205 (2020.01); G06F 40/253 (2020.01)] 63 Claims
OG exemplary drawing
 
1. A method comprising:
receiving, by a chatbot system, an utterance generated by a user interacting with the chatbot system;
inputting, by the chatbot system, the utterance into a machine-learning model comprising a series of network layers, wherein a final network layer of the series of network layers includes a logit function that transforms a first probability for a resolvable class into a first real number representative of a first logit value and a second probability for an unresolvable class into a second real number representative of a second logit value;
determining, by the machine-learning model, the first probability for the resolvable class and the second probability for the unresolvable class;
mapping, by the machine-learning model, the first probability for the resolvable class to the first logit value using the logit function, wherein the logit function for mapping the first probability is a logarithm of odds corresponding to the first probability for the resolvable class, the logarithm of odds being weighted by a centroid of a distribution associated with the resolvable class;
mapping, by the machine-learning model, the second probability for the unresolvable class to an enhanced logit value, wherein the enhanced logit value is a third real number determined independently from the logit function used for mapping the first probability, wherein the enhanced logit value includes: (i) a statistical value determined based on a set of logit values generated from a training dataset; (ii) a bounded value that is selected from a range of values defined by a first logarithm of odds corresponding to the second probability for the unresolvable class, the first logarithm of odds being constrained to a range of values by a bounding function and weighted by a centroid of a distribution associated with the unresolvable class; (iii) a weighted value generated by a second logarithm of odds corresponding to the second probability for the unresolvable class, the second logarithm of odds being constrained to the range of values by the bounding function, scaled by a scaling factor, and weighted by the centroid of the distribution associated with the unresolvable class; (iv) a hyperparameter-optimized value that is generated based on hyperparameter tuning of the machine-learning model; or (v) a learned value that is adjusted during training of the machine-learning model; and
classifying, by the chatbot system, the utterance as the resolvable class or the unresolvable class based on the first logit value and the enhanced logit value.