Maintaining dialogue consistency in natural language models
US12393784B1 · kind B1 · utility
Assignee
Inventor
Key dates
| Filing date | Jan 26, 2022 |
| Grant date | Aug 19, 2025 |
| Priority date | — |
| Expiry date | Jan 22, 2043 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG10L15/26
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
In an embodiment, a method includes receiving a query in a natural language dialogue and submitting first information related to the query to at least one response-routing machine learning model that is trained to evaluate an appropriateness of a set of query response modules. The method also includes selecting a plurality of query response modules from the set of query response modules and routing second information related to the query to the plurality of query response modules. The method also includes receiving a plurality of response candidates responsive to the routing. The method also includes retrieving a plurality of dialogue history candidates from a stored record of dialogue events in the natural language dialogue. The method also includes selecting a response from the plurality of response candidates based, at least in part, on the plurality of dialogue history candidates.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.