Kevin J Leach

US Patent:

20210004539, Jan 7, 2021

Filed:

Sep 1, 2020

Appl. No.:

17/009695

Inventors:

- Ann Arbor MI, US
Stefan Larson - Ann Arbor MI, US
Christopher Clarke - Ann Arbor MI, US
Kevin Leach - Ann Arbor MI, US
Jonathan K. Kummerfeld - Ann Arbor MI, US
Parker Hill - Ann Arbor MI, US
Johann Hauswald - Ann Arbor MI, US
Michael A. Laurenzano - Ann Arbor MI, US
Lingjia Tang - Ann Arbor MI, US
Jason Mars - Ann Arbor MI, US

International Classification:

G06F 40/35
G06N 5/04
G06N 20/00
G06F 40/284

Abstract:

Systems and methods for constructing an artificially diverse corpus of training data includes evaluating a corpus of utterance-based training data samples, identifying a slot replacement candidate; deriving distinct skeleton utterances that include the slot replacement candidate, wherein deriving the distinct skeleton utterances includes replacing slots of each of the plurality of distinct utterance training samples with one of a special token and proper slot classification labels; selecting a subset of the distinct skeleton utterances; converting each of the distinct skeleton utterances of the subset back to distinct utterance training samples while still maintaining the special token at a position of the slot replacement candidate; altering a percentage of the distinct utterance training samples with a distinct randomly-generated slot token value at the position of the slot replacement candidate; and constructing the artificially diverse corpus of training samples based on a collection of the percentage of the distinct utterance training samples.

US Patent:

20200257856, Aug 13, 2020

Filed:

Feb 6, 2020

Appl. No.:

16/783604

Inventors:

- Ann Arbor MI, US
Parker Hill - Ann Arbor MI, US
Kevin Leach - Ann Arbor MI, US
Sean Stapleton - Ann Arbor MI, US
Jonathan K. Kummerfeld - Ann Arbor MI, US
Johann Hauswald - Ann Arbor MI, US
Michael A. Laurenzano - Ann Arbor MI, US
Lingjia Tang - Ann Arbor MI, US
Jason Mars - Ann Arbor MI, US

International Classification:

G06F 40/30
G06N 20/00
G06N 7/00
G06F 40/284

Abstract:

Systems and methods for synthesizing training data for multi-intent utterance segmentation include identifying a first corpus of utterances comprising a plurality of distinct single-intent in-domain utterances; identifying a second corpus of utterances comprising a plurality of distinct single-intent out-of-domain utterances; identifying a third corpus comprising a plurality of distinct conjunction terms; forming a multi-intent training corpus comprising synthetic multi-intent utterances, wherein forming each distinct multi-intent utterance includes: selecting a first distinct in-domain utterance from the first corpus of utterances; probabilistically selecting one of a first out-of-domain utterance from the second corpus and a second in-domain utterance from the first corpus; probabilistically selecting or not selecting a distinct conjunction term from the third corpus; and forming a synthetic multi-intent utterance including appending the first in-domain utterance with one of the first out-of-domain utterance from the second corpus of utterances and the second in-domain utterance from the first corpus of utterances.

US Patent:

20200257857, Aug 13, 2020

Filed:

Apr 21, 2020

Appl. No.:

16/854834

Inventors:

- Ann Arbor MI, US
Parker Hill - Ann Arbor MI, US
Kevin Leach - Ann Arbor MI, US
Sean Stapleton - Ann Arbor MI, US
Jonathan K. Kummerfeld - Ann Arbor MI, US
Johann Hauswald - Ann Arbor MI, US
Michael Laurenzano - Ann Arbor MI, US
Lingjia Tang - Ann Arbor MI, US
Jason Mars - Ann Arbor MI, US

International Classification:

G06F 40/30
G06F 40/284
G06N 20/00
G06N 7/00

Abstract:

Systems and methods for synthesizing training data for multi-intent utterance segmentation include identifying a first corpus of utterances comprising a plurality of distinct single-intent in-domain utterances; identifying a second corpus of utterances comprising a plurality of distinct single-intent out-of-domain utterances; identifying a third corpus comprising a plurality of distinct conjunction terms; forming a multi-intent training corpus comprising synthetic multi-intent utterances, wherein forming each distinct multi-intent utterance includes: selecting a first distinct in-domain utterance from the first corpus of utterances; probabilistically selecting one of a first out-of-domain utterance from the second corpus and a second in-domain utterance from the first corpus; probabilistically selecting or not selecting a distinct conjunction term from the third corpus; and forming a synthetic multi-intent utterance including appending the first in-domain utterance with one of the first out-of-domain utterance from the second corpus of utterances and the second in-domain utterance from the first corpus of utterances.