Back to results
US 12,694,458 B2Granted
System and method for artificial intelligence issue spotting
AssigneeQuest for Justice, LLC
InventorsPatrick Forrest, Binh Dang
Life of this patent
Estimating the term…
Abstract
Systems, methods, and computer-readable storage media for issue spotting in legal cases, and more specifically to training and using an Artificial Intelligence (AI) model to identify possible causes of action in a legal dispute. A system can receive a plurality of cause of action models, and tune an existing AI model using the plurality of action models, resulting in a tuned AI model. The system can then convert the tuned AI model into an AI algorithm, where inputs to the AI algorithm include a jurisdiction and at least one piece of dispute information associated with the dispute, and outputs of the AI algorithm include a cause of action for the dispute and elements of the cause of action.
Claims20
A method comprising:
- training a general use generative Artificial Intelligence (AI) model, the general use generative AI model configured to receive input text and output generative text;
- receiving, at a computer system, a plurality of cause of action records, each record in the cause of action records recording aspects of a historical legal event and containing a cause of action description of the historical legal event;
- executing, via at least one processor of the computer system for each cause of action record in the plurality of cause of action records, a machine-executed search of one or more legal data sources associated with an individual cause of action, resulting in jurisdictional rules for each cause of action record in the plurality of cause of action records;
- preparing, via the at least one processor using the jurisdictional rules, a plurality of cause of action models, wherein each cause of action model in the plurality of cause of action models comprises a cause of action record in the plurality of cause of action records modified according to the jurisdictional rules;
- tuning, via at least one processor of the computer system, the general use generative AI model using the plurality of cause of action models, resulting in a tuned legal AI model;
- and
- converting, via the at least one processor, the tuned legal AI model into an AI algorithm, wherein:
- input to the AI algorithm comprises a jurisdiction and at least one piece of dispute information associated with a dispute;
- and
- output of the AI algorithm comprises a cause of action for the dispute and elements of the cause of action.
7 more elements
A system comprising:
- at least one processor;
- and
- a non-transitory computer-readable storage medium having instructions stored which, when executed by the at least one processor, cause the at least one processor to perform operations comprising:
- training a general use generative Artificial Intelligence (AI) model, the general use generative AI model configured to receive input text and output generative text;
- receiving a plurality of cause of action records, each record in the cause of action records recording aspects of a historical legal event and containing a cause of action description of the historical legal event;
- executing, for each cause of action record in the plurality of cause of action records, a machine-executed search of one or more legal data sources associated with an individual cause of action, resulting in jurisdictional rules for each cause of action record in the plurality of cause of action records;
- preparing, using the jurisdictional rules, a plurality of cause of action models, wherein each cause of action model in the plurality of cause of action models comprises a cause of action record in the plurality of cause of action records modified according to the jurisdictional rules;
- tuning the general use generative AI model using the plurality of cause of action models, resulting in a tuned legal AI model;
- and
- converting the tuned legal AI model into an AI algorithm, wherein:
- input to the AI algorithm comprises a jurisdiction and at least one piece of dispute information associated with a dispute;
- and
- output of the AI algorithm comprises a cause of action for the dispute and elements of the cause of action.
10 more elements
A non-transitory computer-readable storage medium having instructions stored which, when executed by at least one processor, cause the at least one processor to perform operations comprising:
- training a general use generative Artificial Intelligence (AI) model, the general use generative AI model configured to receive input text and output generative text;
- receiving a plurality of cause of action records, each record in the cause of action records recording aspects of a historical legal event and containing a cause of action description of the historical legal event;
- executing, for each cause of action record in the plurality of cause of action records, a machine-executed search of one or more legal data sources associated with an individual cause of action, resulting in jurisdictional rules for each cause of action record in the plurality of cause of action records;
- preparing, using the jurisdictional rules, a plurality of cause of action models, wherein each cause of action model in the plurality of cause of action models comprises a cause of action record in the plurality of cause of action records modified according to the jurisdictional rules;
- tuning the general use generative AI model using the plurality of cause of action models, resulting in a tuned legal AI model;
- and
- converting the tuned AI model into an AI algorithm, wherein:
- input to the AI algorithm comprises a jurisdiction and at least one piece of dispute information associated with a dispute;
- and
- output of the AI algorithm comprises a cause of action for the dispute and elements of the cause of action.
7 more elements
Dependencies are inferred from the claim text.