Granted Patent B2

US 12,470,580 · App. 18/641,610 · Granted Nov 11, 2025

Inventors: Ming-Jung Seow (The Woodlands, TX); Wesley Kenneth Cobb (The Woodlands, TX); Gang Xu (Katy, TX); Tao Yang (Katy, TX); Aaron Poffenberger (Houston, TX); Lon W. Risinger (Katy, TX); Kishor Adinath Saitwal (Houston, TX); Michael S. Yantosca (Katy, TX); David M. Solum (Houston, TX); Alex David Hemsath (Houston, TX); Dennis G. Urech (Katy, TX); Duy Trong Nguyen (Friendswood, TX); Charles Richard Morgan (Katy, TX)

Assignee: Intellective Ai, Inc.

H04L63/1425G06F40/226G06F40/242G06F40/289G06F40/30G06N20/00H04L63/1408G06F40/247G06F40/253G06F40/284G06F40/40

Abstract

Claims (40)

1 . A method, comprising:

initializing, via a processor of a behavior recognition system and using an application programming interface (API) call, a data collection procedure;

identifying, via the processor, data associated with a received request to add the data to a machine learning engine of the behavior recognition system;

retrieving the data via the processor and in response to the received request;

generating, via the processor, a stable model of words based on a statistical distribution of combinations of symbols, the symbols being associated with a normalized version of the data;

calculating, via the processor, a score for at least one word from the stable model of words; and

causing transmission of a behavioral alert based on the score.

2 . The method of claim 1 , wherein the retrieving of the data is from one of a virtual memory file system or a key-value storage device.

3 . The method of claim 1 , wherein the generating the stable model of words is performed by at least one of clustering data from the normalized version of the data or identifying recurring subsets of data from the normalized version of the data, using the machine learning engine.

4 . The method of claim 1 , wherein the generating the stable model of words is performed using the machine learning engine.

5 . The method of claim 1 , further comprising receiving an ordered stream of data, the symbols further associated with clusters generated from the received ordered stream of data.

6 . The method of claim 1 , wherein each word from the stable model of words represents an activity from a plurality of activities.

7 . The method of claim 1 , further comprising generating a neuro-linguistic model based on the stable model of words, the neuro-linguistic model configured to recognize recurring events in a video stream.

8 . A non-transitory, processor-readable medium storing instructions that, when executed, cause a processor of a behavior recognition system to:

initialize, using an application programming interface (API) call, a data collector of the behavior recognition system;

identify data associated with a received request to add the data to a machine learning engine of the behavior recognition system;

retrieve the data via the processor and in response to the received request;

generate a stable model of words based on a statistical distribution of combinations of symbols, the symbols being associated with a normalized version of the data;

calculate a score for at least one word from the stable model of words; and

cause transmission of a behavioral alert based on the score.

9 . The non-transitory, processor-readable medium of claim 8 , further storing instructions to cause the processor to generate the normalized version of the data using the data collector.

10 . The non-transitory, processor-readable medium of claim 8 , further storing instructions to maintain a historical buffer storing transaction data for each data collector from a plurality of data collectors including the data collector.

11 . The non-transitory, processor-readable medium of claim 8 , wherein the data includes sensor data, the memory further storing instructions to cause the processor to aggregate the sensor data in a time-sorted manner.

12 . The non-transitory, processor-readable medium of claim 8 , further storing instructions to cause the processor to receive an ordered stream of data, the symbols being associated with clusters generated from the received ordered stream of data.

13 . The non-transitory, processor-readable medium of claim 8 , wherein each word from the stable model of words represents an activity from a plurality of activities.

14 . The non-transitory, processor-readable medium of claim 8 , further storing instructions to cause the processor to generate a neuro-linguistic model based on the stable model of words, the neuro-linguistic model configured to recognize recurring events in a data stream.

15 . A system, comprising:

a processor; and

a memory storing instructions to cause the processor to:

initialize, using an application programming interface (API) call, a data collection procedure;

identify data associated with a received request to add the data to a machine learning engine of the system;

retrieve the data via the processor and in response to the received request;

generate a stable model of words based on a statistical distribution of combinations of symbols, the symbols being associated with a normalized version of the data;

calculate a plurality of scores for the stable model of words; and

cause transmission of behavioral alerts based on the plurality of scores.

16 . The system of claim 15 , wherein the symbols include fuzzy representations of data from the normalized version of the data.

17 . The system of claim 15 , wherein the instructions to retrieve the data include instructions to retrieve the data from at least one of a virtual memory file system or a key-value storage device.

18 . The system of claim 15 , wherein the instructions to generate the stable model of words include instructions to generate the stable model of words by at least one of clustering data from the normalized version of the data or identifying recurring subsets of data from the normalized version of the data, using the machine learning engine.

19 . The system of claim 15 , further comprising a plurality of data stores communicatively coupled to the machine learning engine and storing at least one of sensor properties, server properties, driver properties, or properties of the system.

20 . The system of claim 19 , wherein the plurality of data stores includes at least one NoSQL database.

Assignments (5)

Continuity (6)

References Cited (246)