US 11,657,810 B2
Query routing for bot-based query response
Steven Ware Jones, Queens, NY (US); Jacob Lewis, New York City, NY (US); Shuai Wang, Sunnyvale, CA (US); Jennifer A. Mallette, Vienna, VA (US); Ruchi Asthana, New York, NY (US); Jia Liu, Astria, NY (US); and Vivek Salve, Poughkeepsie, NY (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Jul. 27, 2020, as Appl. No. 16/939,520.
Prior Publication US 2022/0028378 A1, Jan. 27, 2022
Int. Cl. G10L 15/22 (2006.01); G10L 15/08 (2006.01); G10L 15/30 (2013.01); G06F 16/9532 (2019.01); G06F 16/9538 (2019.01); G06F 16/951 (2019.01)
CPC G10L 15/22 (2013.01) [G06F 16/951 (2019.01); G06F 16/9532 (2019.01); G06F 16/9538 (2019.01); G10L 15/08 (2013.01); G10L 15/30 (2013.01); G10L 2015/221 (2013.01); G10L 2015/227 (2013.01)] 14 Claims
OG exemplary drawing
 
1. A computer-implemented method, comprising:
receiving a query from a user;
analyzing the query to identify a set of entities associated with the query and generate an utterance representing the query;
identifying a set of values associated with a subset of entities of the set of entities, in response to identifying the set of values;
identifying query entities from the subset of entities;
passing the utterance to a plurality of intent classifiers;
generating an intent classification for the utterance based on the plurality of intent classifiers and a comparison of the set of entities with entities associated with the plurality of intent classifiers,
each intent classifier comparing the query entities of the utterance with one or more entities associated with the intent classifier, and
each intent classifier generating a confidence score for the utterance based on comparing the query entities with the one or more entities associated with the intent classifier;
generating a vector for the query based on the set of entities, the utterance, and the intent classification;
determining an answer resource for the query based on the vector and the intent classification of the query; and
in response to determining the answer resource, providing an answer interface based on the query, the vector, and the intent classification such that the answer interface dynamically provides a response to the query.