Method and system for identification of relation among rule intents from a document

1. A processor-implemented method for identification of relation among rule intents from a document, the method comprising:

extracting a plurality of rule sentences from the documents as an input;

mining rule intents from the extracted plurality of rule sentences;

identifying pair-wise relations between the mined rule intents using a trained classifier;

creating a plurality of graphs based on the mined rule intents and the identified pair-wise relations;

optimizing the created plurality of graphs individually to result in a single node based on a plurality of relation optimizing heuristic rules, wherein optimization includes pruning and merging of nodes and edges the graph based on plurality of heuristic rules is performed in several stages that include optimization of a amod node, a satellite node and a triangle node, wherein the amod node is an argument pair-wise relation node, the satellite node is a node of the graph that has only one neighbor and the triangle node is a node of the graph that has three neighbors; and

displaying the single node in a Semantics of Business Vocabulary and Rules (SBVR) format, wherein the single node represents relation among the rule intents.

2. The method of claim 1 , wherein the documents comprises of the plurality of input sentences, the documents are structured or unstructured documents written in natural language and contains noise.

3. The method of claim 1 , wherein the rule intents which are basic atomic facts present in a sentence are mined from extracted rule sentences based on a plurality of rule intent mining heuristic rules.

4. The method of claim 1 , wherein the mined rule intents are saved as vocabularies in the Semantics of Business Vocabulary and Rules (SBVR) format.

5. The method of claim 1 , wherein a classifier is trained and tested to identify pair-wise relations based on matching and feature extraction of annotated extracted rule intents and relation pairs.

6. The method of claim 1 , further comprising the step of identifying a plurality of pairwise relations that include and, or, argument, implication and null using the trained classifier.

7. The method of claim 1 , wherein the graph is created using extracted pair-wise relations as edges and rule intents as nodes.

8. A system comprising:

a rule relation mining unit ( 102 ) for mining rule intents from documents;

a memory ( 104 ) for storing instructions;

one or more communication interfaces ( 106 );

one or more hardware processors ( 108 ) communicatively coupled to the memory using the one or more communication interfaces, wherein the one or more hardware processors are configured by the instructions to execute the one or more modules of the relation mining unit ( 102 ) comprising:

a rule sentence extractor ( 202 ) for extracting a plurality of sentences from the documents as an input;

a rule intent mining unit ( 204 ) for mining rule intents from the extracted plurality of rule sentences;

a classifier ( 208 ) for identifying pair-wise relations between the mined rule intents;

a graph creation unit ( 210 ) for creating a plurality of graphs based on the mined rule intents and the identified pair-wise relations;

a relation optimizer ( 212 ) for optimizing the created plurality of graphs individually to result in a single node based on a plurality of relation optimizing heuristic rules, wherein optimization includes pruning and merging of nodes and edges the graph based on plurality of heuristic rules is performed in several stages that include optimization of a amod node, a satellite node and a triangle node, wherein the amod node is an argument pair-wise relation node, the satellite node is a node of the graph that has only one neighbor and the triangle node is a node of the graph that has three neighbors; and

a display module ( 214 ) for displaying the single node in a Semantics of Business Vocabulary and Rules (SBVR) format, wherein the single node represents relation among the rule intents.

9. The system of claim 8 , wherein the mined rule intents are saved as vocabularies in the Semantics of Business Vocabulary and Rules (SBVR) format in a rule-base database ( 206 ).

10. One or more non-transitory machine readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause:

extracting a plurality of rule sentences from the documents as an input;

mining rule intents from the extracted plurality of rule sentences;

identifying pair-wise relations between the mined rule intents using a trained classifier;

creating a plurality of graphs based on the mined rule intents and the extracted pair-wise relations;

optimizing the created plurality of graphs individually to result in a single node based on a plurality of relation optimizing heuristic rules, wherein optimization includes pruning and merging of nodes and edges the graph based on plurality of heuristic rules is performed in several stages that include optimization of a amod node, a satellite node and a triangle node, wherein the amod node is an argument pair-wise relation node, the satellite node is a node of the graph that has only one neighbor and the triangle node is a node of the graph that has three neighbors; and

displaying the single node in a Semantics of Business Vocabulary and Rules (SBVR) format, wherein the single node represents relation among the rule intents.