US 12,585,643 B2Granted

Generated content source attribution

AssigneeSalesforce, Inc.

InventorDustin Allen Kirk

Life of this patent

Estimating the term…

Abstract

A large language model (LLM) may receive a query via a user interface of a client device. The LLM may generate one or more data queries from the query to query one or more data sets. The LLM may then transmit the one or more data queries to the one or more data sets. The LLM may then receive information associated with the query and a source for the information from the one or more data sets. The source may be indicative of a location within the one or more data sets from where the information was obtained. Following, the LLM may generate a response to the query that includes the information associated with the query and the source for the information and transmit the response to the user interface of the client device for display.

Claims17

Dependent claims

A method for data processing, comprising:

receiving, via a user interface, a natural language query;
parsing, via a large language model that is trained using a plurality of data sets, the natural language query into one or more actionable queries, wherein each one of the one or more actionable queries comprises an indication of a respective data set of the plurality of data sets and an indication of respective information to obtain from the respective data set;
generating, via the large language model and utilizing the one or more actionable queries, one or more data queries from the natural language query, generation of a respective data query of the one or more data queries comprising transforming a respective actionable query into the respective data query that is capable of interacting with the respective data set indicated via the respective actionable query to obtain the respective information included in the respective actionable query, wherein the generation of the one or more data queries to interact with respective data sets is based at least in part on the large language model being trained using the plurality of data sets;
transmitting the one or more data queries to one or more data sets of the plurality of data sets;
receiving, from the one or more data sets of the plurality of data sets and in response to the one or more data queries, information associated with the natural language query and a source for the information, the source being indicative of a location within the one or more data sets from which the information is obtained;
generating, via the large language model, a response to the natural language query, the response comprising an inference generated by the large language model that is based at least in part on the information associated with the natural language query obtained from the one or more data sets and the source for the information that the inference is based on, wherein the response is generated to comprise the source for the information that the inference is based on based at least in part on the large language model that generates the response being trained using the plurality of data sets;
and
transmitting, to the user interface for display, the response to the natural language query that comprises the inference of the information and the source for the information that the inference is based on.

5 more elements

An apparatus for data processing, comprising:

one or more memories storing processor-executable code;
and
one or more processors coupled with the one or more memories and individually or collectively operable to execute the code to cause the apparatus to:
receive, via a user interface, a natural language query;
parse, via a large language model that is trained using a plurality of data sets, the natural language query into one or more actionable queries, wherein each one of the one or more actionable queries comprises an indication of a respective data set of the plurality of data sets and an indication of respective information to obtain from the respective data set;
generate, via the large language model and utilizing the one or more actionable queries, one or more data queries from the natural language query, generation of a respective data query of the one or more data queries comprising transforming a respective actionable query into the respective data query that is capable of interacting with the respective data set indicated via the respective actionable query to obtain the respective information included in the respective actionable query, wherein the generation of the one or more data queries to interact with respective data sets is based at least in part on the large language model being trained using the plurality of data sets;
transmit the one or more data queries to one or more data sets of the plurality of data sets;
receive, from the one or more data sets of the plurality of data sets and in response to the one or more data queries, information associated with the natural language query and a source for the information, the source being indicative of a location within the one or more data sets from which the information is obtained;
generate, via the large language model, a response to the natural language query, the response comprising an inference generated by the large language model that is based at least in part on the information associated with the natural language query obtained from the one or more data sets and the source for the information that the inference is based on, wherein the response is generated to comprise the source for the information that the inference is based on based at least in part on the large language model that generates the response being trained using the plurality of data sets;
and
transmit, to the user interface for display, the response to the natural language query that comprises the inference of the information and the source for the information that the inference is based on.

8 more elements

A non-transitory computer-readable medium storing code for data processing, the code comprising instructions executable by one or more processors to:

receive, via a user interface, a natural language query;
parse, via a large language model that is trained using a plurality of data sets, the natural language query into one or more actionable queries, wherein each one of the one or more actionable queries comprises an indication of a respective data set of the plurality of data sets and an indication of respective information to obtain from the respective data set;
generate, via the large language model and utilizing the one or more actionable queries, one or more data queries from the natural language query, generation of a respective data query of the one or more data queries comprising transforming a respective actionable query into the respective data query that is capable of interacting with the respective data set indicated via the respective actionable query to obtain the respective information included in the respective actionable query, wherein the generation of the one or more data queries to interact with respective data sets is based at least in part on the large language model being trained using the plurality of data sets;
transmit the one or more data queries to one or more data sets of the plurality of data sets;
receive, from the one or more data sets of the plurality of data sets and in response to the one or more data queries, information associated with the natural language query and a source for the information, the source being indicative of a location within the one or more data sets from which the information is obtained;
generate, via the large language model, a response to the natural language query, the response comprising an inference generated by the large language model that is based at least in part on the information associated with the natural language query obtained from the one or more data sets and the source for the information that the inference is based on, wherein the response is generated to comprise the source for the information that the inference is based on based at least in part on the large language model that generates the response being trained using the plurality of data sets;
and
transmit, to the user interface for display, the response to the natural language query that comprises the inference of the information and the source for the information that the inference is based on.

5 more elements

Dependencies are inferred from the claim text.

patentgrep

Loading patent…

Abstract

A method for data processing, comprising:

receiving, via a user interface, a natural language query;
parsing, via a large language model that is trained using a plurality of data sets, the natural language query into one or more actionable queries, wherein each one of the one or more actionable queries comprises an indication of a respective data set of the plurality of data sets and an indication of respective information to obtain from the respective data set;
generating, via the large language model and utilizing the one or more actionable queries, one or more data queries from the natural language query, generation of a respective data query of the one or more data queries comprising transforming a respective actionable query into the respective data query that is capable of interacting with the respective data set indicated via the respective actionable query to obtain the respective information included in the respective actionable query, wherein the generation of the one or more data queries to interact with respective data sets is based at least in part on the large language model being trained using the plurality of data sets;
transmitting the one or more data queries to one or more data sets of the plurality of data sets;
receiving, from the one or more data sets of the plurality of data sets and in response to the one or more data queries, information associated with the natural language query and a source for the information, the source being indicative of a location within the one or more data sets from which the information is obtained;
generating, via the large language model, a response to the natural language query, the response comprising an inference generated by the large language model that is based at least in part on the information associated with the natural language query obtained from the one or more data sets and the source for the information that the inference is based on, wherein the response is generated to comprise the source for the information that the inference is based on based at least in part on the large language model that generates the response being trained using the plurality of data sets;
and
transmitting, to the user interface for display, the response to the natural language query that comprises the inference of the information and the source for the information that the inference is based on.

5 more elements