US 11,657,805 B2
Dynamic context-based routing of speech processing
Ponnu Jacob, Seattle, WA (US); Uday Kumar Kollu, Seattle, WA (US); Jingqian Zhao, Bellevue, WA (US); Prathap Ramachandra, Kirkland, WA (US); Adam Kalman, Seattle, WA (US); Ruiqi Luo, Kenmore, WA (US); Krupal Maddipati, Redmond, WA (US); Charlotte Alizerine Dzialo, Seattle, WA (US); Wenbo Yan, Redmond, WA (US); Liu Yang, Seattle, WA (US); Mohammad Alnuaimat, Seattle, WA (US); Meng Xie, Kirkland, WA (US); Nalledath P Vinodkrishnan, Bellevue, WA (US); and Adriano Devillaine, Sammamish, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Jun. 24, 2021, as Appl. No. 17/304,714.
Prior Publication US 2022/0415310 A1, Dec. 29, 2022
Int. Cl. G10L 15/18 (2013.01); H04L 67/306 (2022.01); G10L 15/22 (2006.01); G06F 16/2457 (2019.01); H04L 67/10 (2022.01); H04L 67/63 (2022.01); G10L 15/30 (2013.01)
CPC G10L 15/1815 (2013.01) [G06F 16/24578 (2019.01); G10L 15/22 (2013.01); H04L 67/10 (2013.01); H04L 67/306 (2013.01); H04L 67/63 (2022.05); G10L 15/30 (2013.01); G10L 2015/223 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
as performed by a computing system comprising one or more computer processors configured to execute specific instructions,
receiving a routing confidence request regarding whether any application of a domain is configured to generate a response to an utterance, wherein the routing confidence request is associated with a plurality of semantic representations of the utterance;
obtaining a set of contextual data items associated with the utterance;
identifying at least a subset of routing confidence providers associated with the domain based at least partly on an association of individual routing confidence providers of the subset with individual semantic representations of the plurality of semantic representations;
obtaining a plurality of routing confidence evaluations using the subset of routing confidence providers, the plurality of semantic representations and the set of contextual data items, wherein a routing confidence evaluation represents a degree of confidence that an application is to generate a response to the utterance; and
generating a response to the routing confidence request based at least partly on the plurality of routing confidence evaluations.