Systems and methods to determine and utilize semantic relatedness between multiple natural language sources to determine strengths and weaknesses

This invention describes a computer program that takes messy written text and turns it into organized information about the things mentioned in it. It does this by first breaking the text into small pieces. Then, a computer model learns to identify and combine these pieces in the best way to create a clear, structured representation of the original text's content.

Why it matters: Filed just as transformer models emerged, the specific segmentation and recombination optimization described would now be greatly simplified or superseded by large language models. These models excel at understanding semantic relatedness and extracting structured data from text.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedSpecialized software development, machine learning expertise.

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10