Automated Patent Drafting via ML and Rules

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The profitability of patent preparation for law firms has declined due to market forces and escalating hourly rates, leading to a climate where only entry-level and non-attorney practitioners can be profitable, and there is a talent shortage despite increasing client demand for patent drafting.

Innovation Solution

The implementation of cutting-edge machine learning and natural language generation technologies to automate the generation of patent application drafts, allowing practitioners to focus on client experience and key aspects of patent preparation, with systems configured to provide data structures representing patent claims, modify them into prose, and create patent specifications without human intervention.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If automated generation of patent application drafts is implemented using machine learning and natural language generation technologies, then productivity and turnaround time are improved, but device complexity increases

Engineering Contradiction:

Improvepatent draft generation speedVSAvoidautomation system complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent draft generation process is divided into distinct segments: claim processing, data structure creation, natural language generation, and final assembly. Each segment is handled by specialized modules that process specific aspects of the patent application, allowing for efficient parallel processing while maintaining overall system manageability.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Structured data structures serve as intermediaries between the input claims and the final natural language patent draft. The system converts claims into standardized data representations, then uses these structured intermediates to generate the prose portion, bridging the gap between technical input and linguistic output.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Ease of operation

If practitioners focus on client experience and key aspects of patent preparation, then service quality is improved, but the extent of automation must increase to handle routine tasks

Engineering Contradiction:

Improvepractitioner service qualityVSAvoidautomated task coverage

Core Design Contradiction:

Ease of operationVSExtent of automation

Solution Approach 1:

The system enables self-service generation of patent specifications by automatically processing claims and producing draft content without requiring practitioner intervention for routine drafting tasks. The automated system handles data processing, structure creation, and text generation independently, freeing practitioners to focus on higher-value activities.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system performs preliminary actions by pre-processing claims into structured data formats and pre-generating specification content before practitioner review. This preliminary automation handles the mechanical aspects of draft creation, allowing practitioners to focus on strategic decisions and client communication.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11651160B2Systems and methods for using machine learning and rules-based algorithms to create a patent specification based on human-provided patent claims such that the patent specification is created without human intervention

Publication Date: 2023.05.16 PAXIMAL INC

AI summary

Systems and methods for using machine learning and rules-based algorithms to create a patent specification based on human-provided patent claims such that the patent specification is created without human intervention are disclosed. Exemplary implementations may: obtain a claim set; obtain a first data structure representing the claim set; obtain a second data structure; obtain a third data structure; and determine one or more sections of the patent specification based on the first data structure, the second data structure, and the third data structure.