Patent Claim Generation Using Neural Encoding and Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated systems struggle to draft patent claims accurately and efficiently, lacking standardization and requiring significant human intervention due to the variability in textual information and expertise of patent attorneys.
Innovation Solution
A method and system utilizing convolutional neural networks (CNN) and recurrent neural networks (RNN) to encode and decode patent document images and text, generating claim sets through a trained network architecture, incorporating deep learning and natural language processing techniques to automate the patent drafting process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated systems are used to draft patent claims, then productivity is improved, but manufacturing precision deteriorates due to lack of standardization and accuracy
Solution Approach 1:
The system performs preliminary encoding of patent document images into feature vectors before claim generation, and pre-trains neural networks on large corpora of patent documents to learn drafting patterns and standards in advance, enabling accurate automated claim generation
Solution Approach 2:
The patent introduces intermediate representations (feature vectors from CNN encoding) and trained neural network models as mediators between the input patent document images and the output claims, enabling the system to bridge the gap between image input and text generation with high accuracy
2Manufacturing precision
If manual drafting by patent attorneys is used, then manufacturing precision is improved, but productivity deteriorates due to time-consuming processes
Solution Approach 1:
The system enables self-service automated claim drafting by training neural networks to independently generate claims from patent document images without requiring manual intervention, while maintaining accuracy through learning from extensive training data
Solution Approach 2:
The system changes the parameters of the drafting process by using deep learning models with adjustable weights and biases that are optimized during training, allowing the system to achieve high accuracy while operating automatically at high speed
3Manufacturing precision
If standardized practices are implemented for claim drafting, then manufacturing precision is improved, but adaptability deteriorates due to variability in technological expertise
Solution Approach 1:
The trained neural network model serves as a universal drafting engine that can handle multiple types of patent documents across different technological fields, applying learned standards consistently while adapting to various input formats and content types
Data Source
AI summary
Method and system for drafting a patent application are presented. The method and system include steps of acquiring at least one text input related to a class of documents; encoding the text input via at least one first network; generating a set of vectors via the at least one first network, where the vector corresponds to a partial representation of the text derived from the at least one first network; obtaining a text corpus for the class of documents, where the text corpus is associated with a language model compiled from the class of documents; decoding the set of vectors based on the text corpus via at least one second network; and obtaining natural language based on the decoding.