US 11,669,351 B2
Intelligent assistant for user-interface to provide geographic event information based on a score which depends on text of conversation
Boris Simanovich, West Hartford, CT (US); Michelle M. Buxton, Colchester, CT (US); Hoa Ton-That, Glastonbury, CT (US); and Anastasia Shekhter, South Windsor, CT (US)
Assigned to THE TRAVELERS INDEMNITY COMPANY, Hartford, CT (US)
Filed by THE TRAVELERS INDEMNITY COMPANY, Hartford, CT (US)
Filed on Jun. 25, 2018, as Appl. No. 16/17,176.
Prior Publication US 2019/0391827 A1, Dec. 26, 2019
Int. Cl. G06F 9/451 (2018.01); H04M 3/527 (2006.01); G06Q 40/08 (2012.01); G06F 40/279 (2020.01); G06Q 30/016 (2023.01); G06N 5/025 (2023.01)
CPC G06F 9/453 (2018.02) [G06F 40/279 (2020.01); G06Q 30/016 (2013.01); G06N 5/025 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An information system providing enhanced prediction of information relevant to a conversation between a requestor and a provider based on multiple levels of context, the information system comprising:
a regional database storing regional information, wherein the regional information comprises information describing physical events within a plurality of geographic regions independent of the requestor; and
a local database storing local information, wherein the local information comprises information describing a current state of the requestor in relation to the provider;
a transaction database storing transactional status information, wherein the transactional status information comprises information describing a plurality of transaction types provided by the provider and a plurality of user-interface suggestions corresponding to the plurality of transaction types;
a conversational intelligence module operatively connected to a communication channel between the requestor and the provider, and configured to:
monitor the conversation between the requestor and the provider,
determine metadata of the conversation, and
determine text of the conversation;
a regional intelligence module operatively connected to the conversational intelligence module and configured to determine a regional status of the requestor based on the metadata of the conversation, the text of the conversation, the regional information, and regional classification rules, wherein determining the regional status comprises determining one or more events of the requestor based on the regional information, the text, and a location of the requestor, and wherein determining the regional status further comprises determining confidence scores of the one or more events;
a local intelligence module operatively connected to the conversational intelligence module and the regional intelligence module, and configured to determine a local status of the requestor based on the text of the conversation, the regional status, the local information, and local classification rules; and
a transactional intelligence module operatively connected to the conversational intelligence module, the regional intelligence module, and the local intelligence module, and configured to:
determine one or more suggestions for the provider based on the regional status, the local status, the transactional status information, and transactional classification rules; and
provide the one or more suggestions to a user-interface device of the provider; and
wherein the determining confidence scores of the one or more events includes determining the confidence scores using, at least in part, the text.