Patent · US Active

Identification and logging of conversations using machine learning

US10819667B2 · kind B2 · utility

1Cited by
6References
17Claims
0Family size

Assignee

Inventors

Key dates

Filing dateMar 9, 2018
Grant dateOct 27, 2020
Priority date
Expiry dateMar 9, 2038

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG10L15/26
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

In one embodiment, a processor receives data indicative of a plurality of conversations involving a primary user. The processor identifies a subset of the plurality of conversations that are regarding a particular topic. The processor adds a conversation to the subset based on a match between one or more keywords in the conversation matching a list of keywords associated with the particular topic. The processor uses a machine learning-based model to identify one or more context characteristics of the conversations in the identified subset. The processor updates the subset of conversations by adding at least one of the conversations to the subset based on the at least one conversation having at least one context characteristic identified by the machine learning-based model. The processor provides data indicative of the updated subset of conversations to a user interface for review by the primary user.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.