IP Library › Granted Patent US 11,321,518
Granted Patent B2
US 11,321,518 · App. 17/200,665 · Granted May 3, 2022

Machine learning based document editing engine

Inventors: Debashis Banerjee (Bengaluru, IN); Prasanna Kumar Govindappa (Bangalore, IN); David Herman (Easton, PA); Krishna Hindhupur Vijay Sudheendra (Bangalore, IN); Shruthi Jinadatta (Bangalore, IN); Anilkumar Tambali (Bangalore, IN); Pravinth Ganesan (Bangalore, IN); Amit Saxena (Bangalore, IN); Gaurav Rathi (Bangalore, IN); Balaji Raghunathan (Bengaluru, IN); Hari Babu Krishnan (Bangalore, IN)
Assignee: SAP SE
G06F40/166G06N3/02G06N20/00
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 11,321,518
App. No.
17/200,665
Granted
May 3, 2022
Kind
B2
Abstract

A method for machine learning based document editing is provided. The method may include receiving, from a client, one or more inputs associated with a document. A recommendation to include and/or exclude a clause, a term, and/or a line item from the document may be generated by at least processing the one or more inputs with a machine learning model. The recommendation to include and/or exclude the clause, the term, and/or the line item from the document may be provided to the client. Related systems and articles of manufacture, including computer program products, are also provided.

Claims (34)

1. A system, comprising:

at least one data processor; and

at least one memory storing instructions which, when executed by the at least one data processor, result in operations comprising:

training a machine learning model by at least processing, with the machine learning model, training data including a first matrix corresponding to a first attribute and a second matrix corresponding to a second attribute, the first matrix including one or more values indicating whether one or more contract clauses are present in a document having the first attribute, the second matrix including one or more values indicating whether the one or more contract clauses are present in a document having the second attribute, and the machine learning model being trained, based at least on the first matrix and the second matrix, to identify one or more correlations between the first attribute, the second attribute, and a presence of the one or more contract clauses;

receiving, from a client, one or more inputs associated with a first document;

generating, by at least processing the one or more inputs with the trained machine learning model, a recommendation to include a first contract clause from the first document, the trained machine learning model generating the recommendation based at least on the first contract clause being included in a second document having one or more attributes and/or content in common with the first document; and

providing, to the client, the recommendation to exclude the first contract clause from the first document.

2. The system of claim 1 , wherein the one or more inputs includes an attribute associated with the first document and/or a second contract clause included in the first document.

3. The system of claim 2 , wherein the first contract clause and/or the second contract clause comprise structured data.

4. The system of claim 1 , wherein the training data further includes a third matrix including one or more values indicating whether each of the one or more contract clauses are present given one or more external factors.

5. The system of claim 4 , wherein the one or more external factors include a current event, a market data, and/or a government regulation.

6. The system of claim 4 , wherein the machine learning model is trained to identify at least one correlation between the second contract clause included in the second document, the second document having the first attribute and/or the second attribute, and/or the external factor.

7. The system of claim 6 , wherein the machine learning model includes at least one kernel, wherein the at least one kernel is associated with one or more weights and biases, and wherein the at least one kernel is configured to identify the at least one correlation by at least applying the one or more weights and biases to the first matrix and/or the second matrix.

8. The system of claim 7 , wherein the training of the machine learning model comprises adjusting the one or more weights and/or biases associated with the at least one kernel, and wherein the one or more weights and/or biases are adjusted to at least minimize an error in the identification of the at least one correlation between the second contract clause included in the second document, the attribute of the second document, and/or the external factor.

9. The system of claim 8 , wherein the adjustment of the one or more weights and/or biases comprises a backward propagation of errors and/or gradient descent.

10. The system of claim 1 , wherein the machine learning model comprises a regression model, an instance-based model, a regularization model, a decision tree, a Bayesian model, a clustering model, an associative model, a neural network, a deep learning model, a dimensionality reduction model, and/or an ensemble model.

11. A computer-implemented method, comprising:

training a machine learning model by at least processing, with the machine learning model, training data including a first matrix corresponding to a first attribute and a second matrix corresponding to a second attribute, the first matrix including one or more values indicating whether one or more contract clauses are present in a document having the first attribute, the second matrix including one or more values indicating whether the one or more contract clauses are present in a document having the second attribute, and the machine learning model being trained, based at least on the first matrix and the second matrix, to identify one or more correlations between the first attribute, the second attribute, and a presence of the one or more contract clauses;

receiving, from a client, one or more inputs associated with a first document;

generating, by at least processing the one or more inputs with the trained machine learning model, a recommendation to include a first contract clause from the first document, the trained machine learning model generating the recommendation based at least on the first contract clause being included in a second document having one or more attributes and/or content in common with the first document; and

providing, to the client, the recommendation to exclude the first contract clause from the first document.

12. The computer-implemented method of claim 11 , wherein the one or more inputs includes an attribute associated with the first document and/or a second contract clause included in the first document.

13. The computer-implemented method of claim 12 , wherein the first contract clause and/or the second contract clause comprise structured data.

14. The computer-implemented method of claim 11 , wherein the training data further includes a third matrix including one or more values indicating whether each of the one or more contract clauses are present and/or absent given one or more external factors.

15. The computer-implemented method of claim 14 , wherein the one or more external factors include a current event, a market data, and/or a government regulation.

16. The computer-implemented method of claim 14 , wherein the machine learning model is trained to identify at least one correlation between the second contract clause included in the second document, the second document having the first attribute and/or the second attribute, and/or the external factor.

17. The computer-implemented method of claim 16 , wherein the machine learning model includes at least one kernel, wherein the at least one kernel is associated with one or more weights and biases, and wherein the at least one kernel is configured to identify the at least one correlation by at least applying the one or more weights and biases to the first matrix and/or the second matrix.

18. The computer-implemented method of claim 17 , wherein the training of the machine learning model comprises adjusting the one or more weights and/or biases associated with the at least one kernel, and wherein the one or more weights and/or biases are adjusted to at least minimize an error in the identification of the at least one correlation between the second contract clause included in the second document, the attribute of the second document, and/or the external factor.

19. The computer-implemented method of claim 18 , wherein the adjustment of the one or more weights and/or biases comprises a backward propagation of errors and/or gradient descent.

20. A non-transitory computer readable medium storing instructions, which when executed by at least one data processor, result in operations comprising:

training a machine learning model by at least processing, with the machine learning model, training data including a first matrix corresponding to a first attribute and a second matrix corresponding to a second attribute, the first matrix including one or more values indicating whether one or more contract clauses are present in a document having the first attribute, the second matrix including one or more values indicating whether the one or more contract clauses are present in a document having the second attribute, and the machine learning model being trained, based at least on the first matrix and the second matrix, to identify one or more correlations between the first attribute, the second attribute, and a presence of the one or more contract clauses;

receiving, from a client, one or more inputs associated with a first document;

generating, by at least processing the one or more inputs with the trained machine learning model, a recommendation to include a first contract clause from the first document, the trained machine learning model generating the recommendation based at least on the first contract clause being included in a second document having one or more attributes and/or content in common with the first document; and

providing, to the client, the recommendation to exclude the first contract clause from the first document.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Mar 16, 2021
From: BANERJEE, DEBASHIS; GOVINDAPPA, PRASANNA KUMAR; HERMAN, DAVID; SUDHEENDRA, KRISHNA HINDHUPUR VIJAY; JINADATTA, SHRUTHI; TAMBALI, ANILKUMAR; GANESAN, PRAVINTH; SAXENA, AMIT; RATHI, GAURAV; RAGHUNATHAN, BALAJI; KRISHNAN, HARI BABU
To: SAP SE
Reel/Frame 055611/0049 →
Continuity (2)
Continuation 15907237 · Feb 27, 2018
Related Publication 20210200936A1 · Jul 1, 2021
Cited By (1)
US 12,387,200