Identifying regulatory data corresponding to executable rules
Various embodiments are provided for correlating regulatory data in a computing environment by a processor. A rule may be associated with one or more textual paragraphs extracted from a policy document that describes at least a portion of the rule.
1 . A method, comprising:
extracting, using natural language processing associated with a neural network, a first set of entities from one or more textual paragraphs associated with a natural language policy document;
converting one or more executable rules, associated with a rule document, into a second set of entities, wherein
each of the one or more executable rules is understandable and executable by a specific computer system;
assigning, each entity of the second set of entities, a ranked score corresponding to a level of importance of each entity of the second set of entities;
extracting, based on the first set of entities, one or more logical structures from the one or more textual paragraphs, wherein
each of the one or more logical structures represents fragments of a respective textual paragraph of the one or more textual paragraphs;
extracting, based on the second set of entities, one or more rule logical structures from the one or more executable rules, wherein
each of the one or more rule logical structures represents at least one of: a mathematical operator, a logical operator, a variable, or a constant;
comparing the one or more logical structures corresponding to the one or more textual paragraphs with the one or more rule logical structures corresponding to the one or more executable rules; and
identifying, for each executable rule of the one or more executable rules, using the natural language processing, a textual paragraph of the one or more textual paragraphs, based on the assigning of the ranked score and the comparing of the one or more logical structures with the one or more rule logical structures, wherein
the textual paragraph corresponds to a highest matching score, among the one or more textual paragraphs, that correlates to an executable rule of the one or more executable rules.
2 . The method of claim 1 , further comprising mapping, using a forward-backward translation operation, the one or more executable rules to the one or more textual paragraphs.
3 . The method of claim 1 , wherein the identifying of the textual paragraph of the one or more textual paragraphs further comprises:
assigning a set of matching scores between the one or more textual paragraphs and the one or more executable rules, wherein
each matching score of the set of matching scores corresponds to a respective textual paragraph of the one or more textual paragraphs and a respective rule of the one or more executable rules,
the set of matching scores includes the highest matching score, and
the set of matching scores indicates a degree of corresponding relevance between the one or more textual paragraphs and the one or more executable rules.
4 . The method of claim 1 , further comprising assigning a confidence score to the textual paragraph of the one or more textual paragraphs, wherein
the confidence score of the textual paragraph of the one or more textual paragraphs indicates a degree of confidence that the textual paragraph of the one or more textual paragraphs matches to the executable rule of the one or more executable rules.
5 . A system, comprising:
one or more processors with executable instructions that when executed cause the system to:
extract, using natural language processing associated with a neural network, a first set of entities from one or more textual paragraphs associated with a natural language policy document;
convert one or more executable rules, associated with a rule document, into a second set of entities, wherein
each of the one or more executable rules is understandable and executable by a specific computer system;
assign, each entity of the second set of entities, a ranked score corresponding to a level of importance of each entity of the second set of entities;
extract, based on the first set of entities, one or more logical structures from the one or more textual paragraphs, wherein
each of the one or more logical structures represents fragments of a respective textual paragraph of the one or more textual paragraphs;
extract, based on the second set of entities, one or more rule logical structures from the one or more executable rules, wherein
each of the one or more rule logical structures represents at least one of: a mathematical operator, a logical operator, a variable, or a constant;
compare the one or more logical structures corresponding to the one or more textual paragraphs with the one or more rule logical structures corresponding to the one or more executable rules; and
identify, for each executable rule of the one or more executable rules, using the natural language processing, a textual paragraph of the one or more textual paragraphs, based on the assignment of the ranked score and the comparison of the one or more logical structures with the one or more rule logical structures, wherein
the textual paragraph corresponds to a highest matching score, among the one or more textual paragraphs, that correlates to an executable rule of the one or more executable rules.
6 . The system of claim 5 , wherein the executable instructions when executed further cause the system to map, using a forward-backward translation operation, the one or more executable rules to the one or more textual paragraphs.
7 . The system of claim 5 , wherein the identification of the textual paragraph of the one or more textual paragraphs further comprises:
assign a set of matching scores between the one or more textual paragraphs and the one or more executable rules, wherein
each matching score of the set of matching scores corresponds to a respective textual paragraph of the one or more textual paragraphs and a respective rule of the one or more executable rules,
the set of matching scores includes the highest matching score, and
the set of matching scores indicates a degree of corresponding relevance between the one or more textual paragraphs and the one or more executable rules.
8 . The system of claim 5 , wherein the executable instructions when executed further cause the system to assign a confidence score to the textual paragraph of the one or more textual paragraphs, wherein
the confidence score of the textual paragraph of the one or more textual paragraphs indicates a degree of confidence that the textual paragraph of the one or more textual paragraphs matches to the executable rule of the one or more executable rules.
9 . A computer program product, comprising:
one or more computer readable storage media; and
program instructions collectively stored on the one or more computer readable storage media to perform operations comprising:
extracting, using natural language processing associated with a neural network, a first set of entities from one or more textual paragraphs associated with a natural language policy document;
converting one or more executable rules, associated with a rule document, into a second set of entities, wherein
each of the one or more executable rules is understandable and executable by a specific computer system;
assigning, each entity of the second set of entities, a ranked score corresponding to a level of importance of each entity of the second set of entities;
extracting, based on the first set of entities, one or more logical structures from the one or more textual paragraphs, wherein
each of the one or more logical structures represents fragments of a respective textual paragraph of the one or more textual paragraphs;
extracting, based on the second set of entities, one or more rule logical structures from the one or more executable rules, wherein
each of the one or more rule logical structures represents at least one of: a mathematical operator, a logical operator, a variable, or a constant;
comparing the one or more logical structures corresponding to the one or more textual paragraphs with the one or more rule logical structures corresponding to the one or more executable rules; and
identifying, for each executable rule of the one or more executable rules, using the natural language processing, a textual paragraph of the one or more textual paragraphs, based on the assigning of the ranked score and the comparing of the one or more logical structures with the one or more rule logical structures, wherein
the textual paragraph corresponds to a highest matching score, among the one or more textual paragraphs, that correlates to an executable rule of the one or more executable rules.
10 . The computer program product of claim 9 , wherein the program instructions stored on the one or more computer readable storage media perform the operations further comprising: mapping, using a forward-backward translation operation, the one or more executable rules to the one or more textual paragraphs.
11 . The computer program product of claim 9 , wherein the identifying of the textual paragraph of the one or more textual paragraphs further comprises:
assigning a set of matching scores between the one or more textual paragraphs and the one or more executable rules, wherein
each matching score of the set of matching scores corresponds to a respective textual paragraph of the one or more textual paragraphs and a respective rule of the one or more executable rules,
the set of matching scores includes the highest matching score, and
the set of matching scores indicates a degree of corresponding relevance between the one or more textual paragraphs and the one or more executable rules.
12 . The computer program product of claim 9 , wherein the program instructions stored on the one or more computer readable storage media perform the operations further comprising assigning a confidence score to the textual paragraph of the one or more textual paragraphs, wherein
the confidence score of the textual paragraph of the one or more textual paragraphs indicates a degree of confidence that the textual paragraph of the one or more textual paragraphs matches to the executable rule of the one or more executable rules.