Method for clause generation
A method for generating bespoke clause language for document or individual clause creation using machine learning and natural language processing. By capturing Author usage and preference data an application employing this method can extract, evaluate, compare, and produce clause language meeting an Author's specific requirements both mechanically, the clause appropriately matches the document purpose, and situationally, the clause appropriately matches the tone the Author wishes to convey in the document.
1. A computer implemented method of drafting a new document using language selection determined by the relational classification between language available in a database, and the desired clause language parameters comprising the steps of:
a) collecting user provided desired clause language parameters, either through real-time user interface, or using previously provided captured user language parameters;
b) collecting user provided input to determine a specific document type for generation;
c) analyzing the language available in the database to determine the baseline traits of the available language in the database to determine a baseline traits score;
d) determining the weighting of each clause language parameter based on the user provided desired clause language parameters having the same attribute type category: Identity, Operational, or Qualitative;
e) generating a Suitability Rubric (SR), formed from a 3 axis graph with each axis having a minimum value of zero (0) and a maximum value of one-hundred (100) having an identify axis, an operational axis, and a qualitative axis, wherein each axis' minimum value represents the weighted language parameter value of the language selection without any language matching the parameters and the axis' maximum value represents a combined weighted language parameter value of a language selection containing language with the highest possible amount of language matching the user provided desired clause language parameters;
f) creating a Suitability Rubric Matrix (SRM) by determining the optimum embodiment of desired clause language based on the user provided desired clause language parameters provided by the user, the SRM being mapped on the SR based on the baseline traits score matching the user's desired language parameter
g) identifying relevant clause language from the subset of documents available in the database fitting a required document type parameter;
h) retrieving the relevant clause language from the document subset;
i) mapping each of the retrieved relevant clauses language on to the Suitability Rubric by multiplying the baseline traits score matching the user's desired language parameter by the Suitability Rubric axis weighting determined by the desired language parameters provided by the user;
j) determining the relative distance between each of the retrieved relevant clause language and the SRM by calculating the square root of the sum of the squared difference between each axis value of the 3 axis graph;
k) selecting clause language, meeting a preselected baseline traits score that most closely matches the user provided desired clause language parameters for inclusion for each of the retrieved relevant clause language; and
l) populating the new document with the selected clause language.