Granted Patent B2

US 12,632,745 · App. 17/952,978 · Granted May 19, 2026

Inventor: Simon Tucker (Oakland, CA)

Assignee: Genesee Valley Innovations, LLC

G06N5/022G06N5/025G06Q10/0639G06Q30/0201G06Q30/0202

Abstract

Claims (43)

1 . A computer-implemented system for building context models in real time, comprising:

a database of models for a user, wherein each model comprises actions that represent a different situation;

a server comprising a central processing unit, memory, an input port to receive the models from the database, and an output port, wherein the central processing unit is configured to:

determine a contextual situation of a user;

compare the situation with models each comprising actions that represent a different situation;

determine that none of the models represents the situation of the user;

determine a similarity value for each of at least a portion of the models with the situation of the user;

apply a threshold to the similarity values;

select all models associated with a similarity value that satisfies the threshold;

generate a new model for the situation, to add to the database, by utilizing a weight associated with each of the selected models that carries over to the actions in that model to identify those actions in the selected models for inclusion in the new model; and

select one or more of the actions of the new model for recommended performance by the user.

2 . A system according to claim 1 , wherein the central processing unit collects the contextual data of the user to determine the situation.

3 . A system according to claim 2 , wherein the contextual data is collected via one or more of a cell phone, smart watch, head mounted computing device, and tablet.

4 . A system according to claim 2 , wherein the contextual data comprises one or more of location, time, temperature, acceleration, vital statistics, inventory, movement data, email, text messages, video, and voice recordings.

5 . A system according to claim 2 , wherein the contextual data is collected continuously, periodically, or randomly.

6 . A system according to claim 1 , wherein the central processing unit recommends during the situation the selected actions from the new model to the user.

7 . A system according to claim 6 , wherein the central processing unit selects the actions for recommendation by identifying those actions as most popular or most consistently performed across all the merged models.

8 . A system according to claim 6 , wherein the weight of each table is considered for each action in that table for selecting the one or more actions for recommendation.

9 . A system according to claim 1 , wherein the central processing unit stores the actions of each model in a different table and combines the actions of the new model into a single table.

10 . A system according to claim 9 , wherein the central processing unit represents each table of actions as nodes and edges.

11 . A computer-implemented method for building context models in real time, comprising:

determining a contextual situation of a user;

comparing the situation with models each comprising actions that represent a different situation;

determining that none of the models represents the situation of the user;

determining a similarity value for each of at least a portion of the models with the situation of the user;

applying a threshold to the similarity values;

selecting all models associated with a similarity value that satisfies the threshold;

generating a new model for the situation, to add to the database, by utilizing a weight associated with each of the selected models that carries over to the actions in that model to identify those actions in the selected models for inclusion in the new model; and

selecting one or more of the actions of the new model for recommended performance by the user.

12 . A method according to claim 11 , further comprising:

collecting the contextual data of the user to determine the situation.

13 . A method according to claim 12 , wherein the contextual data is collected via one or more of a cell phone, smart watch, head mounted computing device, and tablet.

14 . A method according to claim 12 , wherein the contextual data comprises one or more of location, time, temperature, acceleration, vital statistics, inventory, movement data, email, text messages, video, and voice recordings.

15 . A method according to claim 12 , wherein the contextual data is collected continuously, periodically, or randomly.

16 . A method according to claim 11 , further comprising:

recommending during the situation the selected actions from the new model to the user.

17 . A method according to claim 16 , wherein the selected actions are identified as most popular or most consistently performed across all the selected models.

18 . A method according to claim 16 , wherein the weight of each model is considered for each action in that table for selecting the one or more actions for recommendation.

19 . A method according to claim 11 , further comprising:

storing the actions of each model in a different table; and

combining the actions of the new model into a single table.

20 . A method according to claim 19 , further comprising:

representing each table of actions as nodes and edges.

Assignments (7)

Continuity (2)

References Cited (63)