System and Method for Generating Legal Documents

• US 20140025608A1
• Filed: 06/19/2013
• Published: 01/23/2014
• Est. Priority Date: 07/20/2012
• Status: Active Grant

• Current Assignee: Iclosings.com Incorporated
• Original Assignee: Iclosings.com Incorporated
• Inventors: Miller, Kenneth Scott, Hall, Adam
• Granted Patent: US 9,355,370 B2
• Time in Patent Office: Days
• US Class Current: 706/12
• CPC Class Codes: G06F 40/186 Templates
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/20 Ensemble learning
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G06Q 50/18 Legal services

Abstract

A system and method for the automated generation of documents for a legal transaction over a network using probabilistic prediction of customary usage. The predictions are generated by user experience, expert rules and machine learned classifiers based on user input of transaction data. The classifiers are constructed and tested on a partitioned dataset consisting of transaction data and legal document and clause selections in previous transactions. In one embodiment, such dataset is collected in a document management system.

Claims

1. A server system for generating legal documents, the server system comprising:

a processor for executing instructions;

a display, operatively coupled to the processor;

an input communications device;

a computer readable medium, operatively coupled to the processor, wherein the computer readable medium contains;

a set of server system instructions that, if executed by the processor, are operable to cause the server system to prompt a first user to initialize the server system with a plurality of legal document forms and a plurality of form selection attributes corresponding to each of the plurality of legal document forms;

a set of learning agent instructions that, if executed by the processor, are operable to cause the server system to construct at least one machine learned, form selection classifier based on a training set, wherein the training set is a first subset of a set of form usage data;

a set of rule engine instructions that, if executed by the processor, are operable to cause the server system to construct a rules engine, the rules engine comprising form selection rules and resources and logic to apply the form selection rules, wherein the form selection rules include the at least one machine learned, form selection classifier; and

a set of document management instructions that, if executed by the processor, are operable to cause the server system to;

prompt a second user to populate the rules engine with transaction data corresponding to the second user'"'"'s legal transaction;

select at least one of the plurality of legal document forms as determined by the populated rules engine;

populate at least one selected legal document form with data corresponding to the transaction data to produce at least one completed legal document; and

display the at least one completed legal document to the second user.

2. The server system of claim 1, wherein the rules engine is configured to perform a comparative evaluation of the form selection rules using a test set, wherein the test set is a second subset of the set of form usage data.

3. The server system of claim 2 wherein the rules engine is configured to assign a reliability rating to each of the form selection rules based on the comparative evaluation of the form selection rules.

4. The server system of claim 3 wherein the set of document management instructions, if executed by the processor, are further operable to cause the server system to:

display a suggested legal document form and its associated reliability rating for the legal transaction, wherein the suggested legal document form and its associated reliability rating are determined by the rules engine; and

add the transaction data and the second user'"'"'s form selection to the set of form usage data.

5. The server system of claim 4, wherein the machine learned, form selection classifier generates probabilistic predictions of form usage derived from the form usage data and the transaction data.

6. The server system of claim 5, wherein at least one of the form selection rules is dynamic, wherein the dynamic form selection rule references the at least one selected legal document.

7. The server system of claim 6 wherein the form selection rules are initialized with expert provided rules.

8. A method for generating legally binding documents, the method comprising:

populating a document management system with a plurality of legal document forms;

populating a rules engine system with a plurality of form selection rules, wherein populating the rules engine system with the plurality of form selection rules further comprises predicting a form usage probability for each of the plurality of legal document forms;

receiving transaction data for a legal transaction, wherein the transaction data includes at least one form selection attribute;

selecting at least one legal document form from the plurality of legal document forms for the legal transaction, wherein selecting the at least one legal document further comprises applying at least one form selection rule to the transaction data;

populating the at least one selected legal document form with data corresponding to the transaction data; and

capturing form usage data associated with the selected legal document.

9. The method of claim 8 wherein predicting the usage probability tbr each of the plurality of legal document forms comprises applying at least one machine learning algorithm to construct at least one form selection classifier with a first subset of the form usage data.

10. The method of claim 9 wherein constructing the at least one form selection classifier further comprises testing the at least one form selection classifier with a test set, wherein the test set is a second subset of the form usage data.

11. The method of claim 10 wherein the form usage data is collected in the document management system.

12. The method of claim 11 wherein populating the rules engine with a plurality of form selection rules further comprises testing the plurality of form selection rules with the test set and assigning a reliability rating to each of the form selection rules based on the results of such testing.

13. The method of claim 12 wherein determining the at least one selected legal document form for the legal transaction further comprises:

displaying at least one suggested legal document form and its associated form selection rules reliability rating.

14. A method for generating legal documents, the method comprising:

populating a document management system with a plurality of legal document forms and a plurality of optional clauses corresponding to at least one of the plurality of legal document forms;

populating a rules engine system with a plurality of form selection rules, wherein populating the rules engine system with a plurality of form selection rules further comprises the rules engine predicting a usage probability for each of the plurality of legal document forms;

populating the rules engine system with a plurality of clause selection rules wherein each clause selection rule maps at least one clause selection attribute to at least one optional clause, wherein populating the rules engine system with a plurality of clause selection rules further comprises the rules engine predicting a usage probability for each of the plurality of optional clauses;

receiving transaction data for a current transaction, wherein the transaction data includes at least one user inputted form selection attribute and at least one user inputted clause selection attribute;

determining with the rules engine at least one selected legal document form for the current legal transaction;

determining with the rules engine at least one selected optional clause for the at least one selected legal document form; and

populating the at least one selected legal document form with the at least one selected optional clause and with data corresponding to the transaction data; and

capturing usage data associated with the selected legal document and the selected optional clause to reiteratively revise usage probabilities.

15. The method of claim 14 wherein predicting the usage probability for each of the plurality of optional clauses comprises applying at least one machine learning algorithm to construct at least one clause selection classifier with a first subset of the clause usage data.

16. The method of claim 15 wherein constructing the at least one clause selection classifier further comprises testing the at least one clause selection classifier using a test set, wherein the test set is a second subset of the clause usage data.

17. The method of claim 16 wherein the set of clause usage data is collected in the document management system.

18. The method of claim 17 wherein populating the rules engine with a plurality of clause selection rules further comprises testing the plurality of clause selection rules with the test set and assigning a reliability rating to each of the clause selection rules based on the test results.

19. The method of claim 18 wherein the at least one selected optional clause contains a plurality of subordinate optional clauses.

20. The method of claim 19 wherein the at least one selected optional clause and the plurality of subordinate optional clauses are organized in a tree structure.