Converting rules in data processing systems

Patent metadata

Field	Value
Publication number	US-10318872-B1
Application number	US-201414210659-A
Country	US
Kind code	B1
Filing date	Mar 14, 2014
Priority date	Mar 15, 2013
Publication date	Jun 11, 2019
Grant date	Jun 11, 2019

Abstract

In one aspect, in general, a computer-implemented method includes receiving an order-dependent list of rules generating a binary decision diagram based on the order-dependent list of rules, the binary decision diagram including decision nodes that include expressions including Boolean operators and operands, and generating an order-independent list of rules based on the binary decision diagram.

First claim

What is claimed is:

1. A computer-implemented method for converting an order-dependent list of data processing rules capable of being processed by a first data processing system comprising one or more computer processors to an order-independent list of data processing rules capable of being processed by a second data processing system comprising one or more computer processors, the method including:
receiving a first order-dependent list of data processing rules,
wherein a rule of the order-dependent list of data processing rules includes a condition that includes a Boolean operator applied to a data variable when the rule is processed by the first data processing system;
identifying the conditions of two of the order-dependent data processing rules of the first order-dependent list of data processing rules;
generating a new order-dependent data processing rule based on the two of the order-dependent data processing rules of the first order-dependent list of data processing rules, the new order-dependent data processing rule including one of the identified conditions from one of the at least two order-dependent data processing rules, and the new order-dependent data processing rule including another identified condition from another of the two order-dependent data processing rules;
generating a second order-dependent list of data processing rules including the new order-dependent data processing rule;
generating a binary decision diagram based on the second order-dependent list of data processing rules, the binary decision diagram including decision nodes that include expressions including Boolean operators and operands;
generating an order-independent list of data processing rules based on the binary decision diagram; and
providing the order-independent list of rules to the second data processing system.

2. The method of claim 1 , in which generating an order-independent list of rules based on the binary decision diagram includes generating a flat expression representation of the binary decision diagram.

3. The method of claim 1 further including:
removing an expression of one of the decision nodes of the binary decision diagram.

4. The method of claim 3 in which removing an expression of one of the decision nodes of the binary decision diagram includes determining a semantic meaning of one of the operands.

5. The method of claim 1 in which one of the decision nodes includes an operand that includes a bit vector.

6. The method of claim 1 in which, when supplied a set of inputs, the first order-dependent list of rules evaluates to a result, and the order-independent list of rules evaluates to the same result.

7. The method of claim 1 in which the first order-dependent list of rules can be processed by a first system but cannot be processed by a second system, the second system being configured to process order-independent rules, and the method includes processing the order-independent list of rules using the second system.

8. The method of claim 1 , wherein at least one rule of the order-dependent list of data processing rules includes at least one condition that includes at least one Boolean operator applied to at least one data variable when the at least one rule is processed by the first data processing system.

9. The method of claim 8 , comprising identifying the conditions of at least two of the order-dependent data processing rules of the first order-dependent list of data processing rules.

10. The method of claim 9 , comprising generating at least one new order-dependent data processing rule based on the at least two of the order-dependent data processing rules of the first order-dependent list of data processing rules, the at least one new order-dependent data processing rule including at least one of the identified conditions from one of the at least two order-dependent data processing rules, and the at least new order-dependent data processing rule including at least one other identified condition from another of the at least two order-dependent data processing rules.

11. The method of claim 10 , comprising generating a second order-dependent list of data processing rules including the at least new order-dependent data processing rule.

12. A computer-implemented method for converting an order-dependent list of data processing rules capable of being processed by a first data processing system comprising one or more computer processors to an order-independent list of data processing rules capable of being processed by a second data processing system comprising one or more computer processors, the method including:
receiving a first set of rules, wherein the first set of rules is based on data processing rules capable of being processed by a first data processing system comprising one or more computer processors, wherein a rule of the first set of rules includes a condition that includes a Boolean operator applied to a data variable when the rule is processed by the first data processing system;
generating a binary decision diagram according to the first set of rules, the binary decision diagram having a first decision node that includes an expression having a Boolean operator and operands, one of the operands including a bit vector having two or more bits, each bit position in the bit vector corresponding to one item in a first list of items, the bit vector representing a second list of items that includes a subset of items selected from the first list of items, the first decision node representing a decision with respect to the second list of items;
generating a second set of rules based on the binary decision diagram; and
providing the second set of rules to a second data processing system comprising one or more computer processors and capable of processing the second set of rules.

13. The method of claim 12 in which each bit position in the bit vector corresponds to a pharmaceutical, and the bit vector represents a list of pharmaceuticals.

14. The method of claim 13 in which the first decision node includes a Boolean operator, the operands include a first operand that represents a first pharmaceutical and a second operand that includes the bit vector, and the expression represents a relation between the first pharmaceutical and the list of pharmaceuticals.

15. The method of claim 12 in which the first and second sets of rules include rules for processing insurance claims.

16. The method of claim 12 in which the first set of rules includes a set of order-dependent rules, and the second set of rules includes a set of order-independent rules.

17. The method of claim 12 in which the binary decision diagram includes a second decision node that includes an expression having a Boolean operator and operands, one of the operands including a bit vector corresponding to the first list of items, and wherein the method includes combining the first and second decision nodes to form a third decision node that includes an operand generated based on a result of a Boolean operation on the bit vector in the first decision node and the bit vector in the second decision node.

18. The method of claim 12 in which, when supplied a set of inputs, the first set of rules evaluates to a result, and the second list of rules evaluates to the same result.

19. A non-transitory computer-readable medium storing software including instructions for causing a computing system to carry out operations for converting an order-dependent list of data processing rules capable of being processed by a first data processing system comprising one or more compute

Assignees

• Ab Initio Technology Llc

Inventors

• Feynman Carl Richard
• Schechter Ian
• Fournier David

Classifications

• G06N5/01

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

• G06N5/025 Primary

  Extracting rules from data · CPC title

• G06N5/04 Primary

  Inference or reasoning models · CPC title