Domain specific intelligence using natural language
AssigneeDisney Enterprises, Inc.
InventorsDavid DeWet RAAL, Jason Alexander COX
Life of this patent
Estimating the term…
Abstract
Techniques for generating a result for a search query. These techniques include identifying a complex search query including more than two parameters, dividing the complex search query into one or more components, and determining an intent for each of the one or more components using machine learning (ML), the ML including at least one of: a large language model (LLM) or natural language processing (NLP) neural network. The techniques further include generating a result for the search query based on routing each component through a pipeline using the respective intent, the pipeline including both a query against a graph database and a search against a vector database.
Claims20
A method, comprising:
- identifying a complex search query comprising more than two parameters;
- dividing the complex search query into one or more components;
- determining a respective intent for each respective component of the one or more components by processing the respective component using machine learning (ML), the ML comprising at least one of:
- a large language model (LLM) or natural language processing (NLP) neural network;
- and
- generating a result for the complex search query, comprising routing each respective component of the one or more components through a pipeline based on the respective intents, comprising, for a first component of the one or more components having a first intent:
- processing the first component using a current step of the pipeline;
- generating a graph database query based at least in part on the current step in the pipeline and the first intent;
- identifying, for the first component, a next step in the pipeline based on executing the graph database query against a knowledge graph;
- and
- processing the first component using the next step.
One or more non-transitory computer readable media containing, in any combination, computer program code that, when executed by operation of any combination of one or more processors, performs operations comprising:
- identifying a complex search query comprising more than two parameters;
- dividing the complex search query into one or more components;
- determining a respective intent for each respective component of the one or more components by processing the respective component using machine learning (ML), the ML comprising at least one of:
- a large language model (LLM) or natural language processing (NLP) neural network;
- and
- generating a result for the complex search query, comprising routing each respective component of the one or more components through a pipeline based on the respective intents, comprising, for a first component of the one or more components having a first intent:
- processing the first component using a current step of the pipeline;
- generating a graph database query based at least in part on the current step in the pipeline and the first intent;
- identifying, for the first component, a next step in the pipeline based on executing the graph database query against a knowledge graph;
- and
- processing the first component using the next step.
A system, comprising:
- one or more processors;
- and
- one or more memories storing a program, which, when executed on any combination of the one or more processors, performs operations, the operations comprising:
- identifying a complex search query comprising more than two parameters;
- dividing the complex search query into one or more components;
- determining a respective intent for each respective component of the one or more components by processing the respective component using machine learning (ML), the ML comprising at least one of:
- a large language model (LLM) or natural language processing (NLP) neural network;
- and
- generating a result for the complex search query, comprising routing each respective component of the one or more components through a pipeline based on the respective intents, comprising, for a first component of the one or more components having a first intent:
- processing the first component using a current step of the pipeline;
- generating a graph database query based at least in part on the current step in the pipeline and the first intent;
- identifying, for the first component, a next step in the pipeline based on executing the graph database query against a knowledge graph;
- and
- processing the first component using the next step.
Dependencies are inferred from the claim text.