Feedback techniques for processing natural language queries
Inventors
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12717826-B2
Publication Date
2026-08-25
Expiration Date
Abstract
Approaches for executing feedback techniques for processing of natural language queries are described. According to one example, upon receiving a customization request in relation to a visual indicator, data fields associated with the visual indicator are displayed. Each of the data fields represents a subset of data stored within one or more datasets, the data being graphically representable through the visual indicator. A first input, including additional synonyms to be assigned to a data field from the data fields, is received. A list of pre-assigned synonyms, associated with the data field for processing of natural language queries for one or more topics, is obtained. The additional synonyms are reconciled with the list of pre-assigned synonyms to obtain an updated list of pre-assigned synonyms for the data field. The updated list of pre-assigned synonyms may be used for future processing of natural language queries raised for the one or more topics.
Core Innovation
The system accepts user-provided synonyms for data fields associated with a visual indicator, compares those synonyms with pre-assigned synonyms, and stores distinct additions in an updated list for future natural-language query processing. The data fields represent subsets of data associated with datasets, and their display and reconciliation are constrained by dataset access and topic authorization. The feedback and pre-assigned synonym lists may be stored in databases managed by different entities.
The supplied content does not include the background section, so a specific problem described there is not explicitly described in patent. It states that the system is intended to improve query-response accuracy while reducing manual updating and unauthorized synonym changes. The updated synonym list is used for future processing of natural-language queries on authorized topics.
Claims Coverage
The patent has three independent claims: a method claim, a system claim, and a non-transitory computer-readable medium claim. Their main inventive features are summarized in three features below.
Reconciliation of user-provided synonyms
The method receives additional synonyms for a data field associated with a visual indicator, obtains the pre-assigned synonyms for that field, reconciles the additional synonyms with the pre-assigned synonyms, and stores the updated list for future processing of natural-language queries.
System for synonym reconciliation and query processing
The system includes a synonym reconciliation module that receives and stores user-provided synonyms, reconciles them with the pre-assigned synonyms for a data field, and stores the updated list. A query processing module uses the updated list to process a natural-language query.
Computer-readable medium for updating synonyms
The non-transitory computer-readable medium comprises instructions to compare additional synonyms with pre-assigned synonyms, determine at least one distinct synonym, incorporate it into an updated list, and store that list for future processing of natural-language queries.
The three independent claims cover the synonym-update process as a method, as a system with query processing, and as executable instructions on a non-transitory computer-readable medium. Each centers on updating pre-assigned synonyms for data fields associated with a visual indicator so the updated list can support natural-language query processing.
Stated Advantages
Improves query-response accuracy.
Reduces manual updating.
Reduces unauthorized synonym changes.
Documented Applications
Processing natural-language queries for authorized topics using updated synonyms associated with data fields.
Generating a response through a visual indicator from natural-language query processing.
Interested in licensing this patent?