Systems and methods for machine learning based rule discovery for data migration

Patent metadata

Field	Value
Publication number	US-12118478-B2
Application number	US-202016998674-A
Country	US
Kind code	B2
Filing date	Aug 20, 2020
Priority date	May 8, 2020
Publication date	Oct 15, 2024
Grant date	Oct 15, 2024

Abstract

Systems and methods for deriving classification rules from documents and a database using rule-based machine learning. The method includes extracting first variables from documents corresponding to an organization. The method further includes extracting second variables from a database corresponding to the organization. The method also includes filtering the extracted second variables based on at least one of null values, repeat variables, location variables, ID variables, or data variables. The method further includes deriving first classification rules based on the first variables using a rule-based machine learning algorithm. The method also includes calculating an accuracy of the derived first classification rules. The method also includes deriving second classification rules based on the first variables and the filtered second variables. The method further includes determining a suggested additional variable based on the derived second classification rules and the calculated accuracy.

First claim

What is claimed:

1. A method for deriving classification rules from documents and a database using rule-based machine learning, the method comprising:
extracting, by a server computing device, a first plurality of variables from documents corresponding to an organization;
extracting, by the server computing device, a second plurality of variables from a database corresponding to the organization;
filtering, by the server computing device, the extracted second plurality of variables based on at least one of null values, repeat variables, location variables, ID variables, or date variables;
deriving, by the server computing device, a first plurality of classification rules based on the first plurality of variables using a rule-based machine learning algorithm, comprising:
a) executing the rule-based machine learning algorithm using the first plurality of variables as input to derive a classification rule and identify a subset of the first plurality of variables that satisfy the classification rule,
b) removing the identified subset of variables from the first plurality of variables,
c) repeating steps a) and b) using the remaining first plurality of variables as input to identify additional subsets of the remaining variables that satisfy additional classification rules, and
d) storing the classification rules derived by the rule-based machine learning algorithm in step a) as the first plurality of classification rules;

calculating, by the server computing device, an accuracy of the derived first plurality of classification rules;
deriving, by the server computing device, a second plurality of classification rules based on the first plurality of variables and filtered second plurality of variables using the rule-based machine learning algorithm, comprising:
e) executing the rule-based machine learning algorithm using the first plurality of variables and one of the filtered second plurality of variables as input to derive a classification rule and identify a subset of the first plurality of variables and one of the filtered second plurality of variables that satisfy the classification rule,
f) removing the identified subset of variables from the first plurality of variables and the one of the filtered second plurality of variables,
g) updating the calculated accuracy of the derived first plurality of classification rules based upon the first plurality of variables and the one of the filtered second plurality of variables,
h) repeating steps e), f) and g) using the remaining first plurality of variables and another one of the filtered second plurality of variables as input to identify additional subsets of the remaining variables that satisfy additional classification rules, and
i) storing the classification rules generated by the rule-based machine learning algorithm in step e) as the second plurality of classification rules; and

identifying, by the server computing device, a suggested additional variable from the filtered second plurality of variables for inclusion in the first plurality of variables based upon the updated accuracy of the derived first plurality of classification rules; and
generating, by the server computing device, for display the derived first plurality of classification rules, the derived second plurality of classification rules, the updated accuracy of the derived first plurality of classification rules, and the suggested additional variable.

2. The method of claim 1 , wherein the server computing device is configured to calculate the accuracy of the derived first plurality of classification rules based on a known plurality of classification rules corresponding to the organization.

3. The method of claim 1 , wherein the server computing device is further configured to extract the first plurality of variables using natural language processing.

4. The method of claim 1 , wherein the database comprises demographic data, employment data, and benefit plan data.

5. The method of claim 1 , wherein the server computing device is further configured to map the extracted first plurality of variables to corresponding entries of the database.

6. The method of claim 1 , wherein the server computing device is further configured to classify each of the extracted first plurality of variables and second plurality of variables as character-based or numeric.

7. The method of claim 1 , wherein the first plurality of classification rules are derived sequentially using the rule-based machine learning algorithm.

8. The method of claim 1 , wherein the server computing device is further configured to sequentially derive the second plurality of classification rules based on the first plurality of variables and the filtered second plurality of variables using the rule-based machine learning algorithm.

9. The method of claim 8 , wherein the server computing device is further configured to calculate an accuracy of the derived second plurality of classification rules based on a known plurality of classification rules corresponding to the organization.

10. A system for deriving classification rules from documents and a database using rule-based machine learning, the system comprising:
a server computing device communicatively coupled to a database corresponding to an organization and a display device, the server computing device configured to:
extract a first plurality of variables from documents corresponding to an organization;
extract a second plurality of variables from the database corresponding to the organization;
filter the extracted second plurality of variables based on at least one of null values, repeat variables, location variables, ID variables, or date variables;
derive a first plurality of classification rules based on the first plurality of variables using a rule-based machine learning algorithm, comprising:
a) executing the rule-based machine learning algorithm using the first plurality of variables as input to derive a classification rule and identify a subset of the first plurality of variables that satisfy the classification rule,
b) removing the identified subset of variables from the first plurality of variables,
c) repeating steps a) and b) using the remaining first plurality of variables as input to identify additional subsets of the remaining variables that satisfy additional classification rules, and
d) storing the classification rules derived by the rule-based machine learning algorithm in step a) as the first plurality of classification rules;

calculate an accuracy of the derived first plurality of classification rules;
derive a second plurality of classification rules based on the first plurality of variables and filtered second plurality of variables using a rule-based machine learning algorithm, comprising:
e) executing the rule-based machine learning algorithm using the first plurality of variables and one of the filtered second plurality of variables as input to derive a classification rule and identify a subset of the first plurality of variables and one of the filtered second plurality of variables that satisfy the classification rule,
f) removing the identified subset of variables from the first plurality of variables and the one of the filtered second plurality of variables,
g) updating the calculated accuracy of the derived first plurality of classification rules based upon the first plurality of variables and the one of the filtered second plurality of variables,
h) repeating steps e), f) and g) using the remaining first plurality of variables and another one of the filtered second plurality of variables as input to identify additional subsets of the remaining variables that satisfy additional classification rules, and
i) storing the classification rules g

Assignees

• Fmr Llc

Inventors

• Divakarmurthy Pramod
• Solomon Hua Zhai
• Raikar Nikhil
• Mani Prabhakar
• Jadhav Amol Vinayak
• Sabir Behlool
• Jia Yugang
• Roy Bidhan
• Homer Mark

Classifications

• G06N20/00

  Machine learning · CPC title

• G06F16/93

  Document management systems · CPC title

• G06F40/279

  Recognition of textual entities · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06N5/025

  Extracting rules from data · CPC title