IP Library Granted Patent US 12,645,762
Granted Patent B2
US 12,645,762 · App. 18/957,583 · Granted Jun 2, 2026

Apparatus and method for textual data analysis and generation of a performance datum

Inventors: Blake Browder (Dallas, TX); Joy Figarsky (Little Rock, AR)
Assignee: BH Operations, LLC
G06F18/2415G06N20/00
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,645,762
App. No.
18/957,583
Granted
Jun 2, 2026
Kind
B2
Abstract

An apparatus and method for textual data analysis and generation of a performance datum are disclosed. The apparatus includes at least a processor and a memory communicatively connected to the at least a processor, wherein the memory contains instructions configuring the at least a processor to receive input data from a plurality of users, wherein the input data comprises textual data, analyze the textual data as a function of referential textual data, identify a pattern datum between the plurality of segments and the plurality of referential segments, generate a performance datum for each user of the plurality of users as a function of the input data, the pattern datum and a plurality of performance metrics and transmit the performance datum to a plurality of user devices.

Claims (48)

1 . An apparatus for textual data analysis and generation of a performance datum, the apparatus comprising:

at least a processor; and

a memory communicatively connected to the at least a processor, wherein the memory contains instructions configuring the at least a processor to:

receive input data from a plurality of users, wherein the input data comprises textual data and includes data aimed at ensuring compliance with safety and regulatory requirements;

analyze the textual data as a function of referential textual data comprising exemplar descriptions associated with an environment in which the textual data was generated, wherein analyzing the textual data comprises:

segmenting the textual data and the referential textual data into a plurality of segments and a plurality of referential segments respectively; and

encoding the plurality of segments and the plurality of referential segments using an encoder, wherein the encoder is configured to convert each segment into a vector representation;

identify a pattern datum between the plurality of segments and the plurality of referential segments by comparing the encoded vector representations of the plurality of segments to the encoded referential vector representations of the plurality of referential segments, wherein identifying the pattern datum comprises:

generating pattern training data, wherein the pattern training data comprises pairs of texts and corresponding similarity labels indicating whether the pairs are semantically similar;

training a pattern machine-learning model using the pattern training data; and

identifying the pattern datum using the trained pattern machine-learning model as a function of an output of the encoder;

generate a performance datum of each user of the plurality of users as a function of the input data, the pattern datum and a plurality of performance metrics comprising a frequency of receipt of the input data, wherein the performance datum comprises a level of completeness of the safety and regulatory requirements; and

transmit the performance datum of each user to a at least a user device.

2 . The apparatus of claim 1 , wherein the referential textual data comprises a second user input.

3 . The apparatus of claim 1 , wherein identifying the pattern datum comprises iteratively updating the pattern training data as a function of the textual data by adding correlations between the textual data and the pattern datum.

4 . The apparatus of claim 1 , wherein identifying the pattern datum comprises generating a prompt to a second user as a function of the pattern datum, wherein the pattern datum comprises dissimilarity between the plurality of segments and the plurality of referential segments.

5 . The apparatus of claim 1 , wherein the plurality of performance metrics comprises a frequency of inputting the input data.

6 . The apparatus of claim 1 , wherein the plurality of performance metrics comprises checklist completeness.

7 . The apparatus of claim 1 , wherein the plurality of performance metrics comprises length of text in the textual data.

8 . The apparatus of claim 1 , wherein generating the performance datum comprises analyzing the input data as a function of the plurality of performance metrics using a natural language processing module, wherein the plurality of performance metrics comprises descriptiveness of texts in the textual data.

9 . The apparatus of claim 1 , wherein generating the performance datum comprises:

generating performance training data, wherein the performance training data comprises exemplary textual data and exemplary pattern data correlated to exemplary performance data;

training a performance machine-learning model using the performance training data; and

generating the performance datum using the trained performance machine-learning model.

10 . The apparatus of claim 9 , wherein generating the performance training data comprises updating the performance training data as a function of an output of the pattern machine-learning model by adding correlations of the pattern datum and the performance datum.

11 . A method for textual data analysis and generation of a performance datum, the method comprising:

receiving, using at least a processor, input data from a plurality of users, wherein the input data comprises textual data and includes data aimed at ensuring compliance with safety and regulatory requirements;

analyzing, using the at least a processor, the textual data as a function of referential textual data comprising exemplar descriptions associated with an environment in which the textual data was generated, wherein analyzing the textual data comprises:

segmenting the textual data and the referential textual data into a plurality of segments and a plurality of referential segments respectively; and

encoding the plurality of segments and the plurality of referential segments using an encoder, wherein the encoder is configured to convert each segment into a vector representation;

identifying, using the at least a processor, a pattern datum between the plurality of segments and the plurality of referential segments by comparing the encoded vector representations of the plurality of segments to the encoded referential vector representations of the plurality of referential segments, wherein identifying the pattern datum comprises:

generating pattern training data, wherein the pattern training data comprises pairs of texts and corresponding similarity labels indicating whether the pairs are semantically similar;

training a pattern machine-learning model using the pattern training data; and

identifying the pattern datum using the trained pattern machine-learning model as a function of an output of the encoder;

generating, using the at least a processor, a performance datum for each user of the plurality of users as a function of the input data, the pattern datum and a plurality of performance metrics comprising a frequency of receipt of the input data, wherein the performance datum comprises a level of completeness of the safety and regulatory requirements; and

transmitting, using the at least a processor, the performance datum to at least a user device.

12 . The method of claim 11 , wherein the referential textual data comprises a second user input.

13 . The method of claim 11 , wherein identifying the pattern datum comprises iteratively updating the pattern training data as a function of the textual data by adding correlations between the textual data and the pattern datum.

14 . The method of claim 11 , wherein identifying the pattern datum comprises generating a prompt to a second user as a function of the pattern datum, wherein the pattern datum comprises dissimilarity between the plurality of segments and the plurality of referential segments.

15 . The method of claim 11 , wherein the plurality of performance metrics comprises a frequency of inputting the input data.

16 . The method of claim 11 , wherein the plurality of performance metrics comprises checklist completeness.

17 . The method of claim 11 , wherein the plurality of performance metrics comprises length of text in the textual data.

18 . The method of claim 11 , wherein generating the performance datum comprises analyzing the input data as a function of the plurality of performance metrics using a natural language processing module, wherein the plurality of performance metrics comprises descriptiveness of texts in the textual data.

19 . The method of claim 11 , wherein generating the performance datum comprises:

generating performance training data, wherein the performance training data comprises exemplary textual data and exemplary pattern data correlated to exemplary performance datums;

training a performance machine-learning model using the performance training data; and

generating the performance datum using the trained performance machine-learning model.

20 . The method of claim 19 , wherein generating the performance training data comprises updating the performance training data as a function of an output of the pattern machine-learning model by adding correlations of the pattern datum and the performance datum.

Assignments (2)
CORRECTIVE ASSIGNMENT TO CORRECT THE ASSIGNEE PREVIOUSLY RECORDED ON REEL 72292 FRAME 767. ASSIGNOR(S) HEREBY CONFIRMS THE ASSIGNMENT. Recorded Dec 16, 2025
From: SIGNET HEALTH CORPORATION
To: BH OPERATIONS, LLC
Reel/Frame 073992/0817 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 31, 2025
From: SIGNET HEALTH CORPORATION
To: BEHAVIORAL HEALTH OPERATIONS, LLC
Reel/Frame 072292/0767 →
Continuity (1)
Related Publication 20260147857A1 · May 28, 2026
References Cited (11)
US 9240120B2 · Girardeau et al. · 2016 [cited by applicant]
US 12046359B2 · Durlach et al. · 2024 [cited by applicant]
US 20080306943A1 · Patterson · 2008 [cited by examiner]
US 20220393875A1 · Shrinivasan · 2022 [cited by examiner]
US 20230334389A1 · Dalley, Jr. et al. · 2023 [cited by applicant]
EP 3276554A1 · 2018 [cited by applicant]
Hunt, Ethan, Ritvik Janamsetty, Chanana Kinares, Chanel Koh, Alexis Sanchez, Felix Zhan, Murat Ozdemir et al. “Machine learning models for paraphrase identification and its applications on plagiarism detection.” In 2019… [cited by examiner]
Somers, Rick, Sam Cunningham, Sarah Dart, Sheona Thomson, Caslon Chua, and Edmund Pickering. “AssignmentWatch: An automated detection and alert tool for reducing academic misconduct associated with file-sharing websites… [cited by examiner]
Foltýnek, Tomáš, Norman Meuschke, and Bela Gipp. “Academic plagiarism detection: a systematic literature review.” ACM Computing Surveys (CSUR) 52, No. 6 (2019): 1-42. (Year: 2019). [cited by examiner]
Kruiper R, Kumar B, Watson R, Sadeghineko F, Gray A, Konstas I. A platform-based Natural Language processing-driven strategy for digitalising regulatory compliance processes for the built environment. Advanced Engineeri… [cited by examiner]
Kulp, Leah, Aleksandra Sarcevic, Yinan Zheng, Megan Cheng, Emily Alberto, and Randall Burd. “Checklist design reconsidered: Understanding checklist compliance and timing of interactions.” In Proceedings of the 2020 CHI … [cited by examiner]