Systems and methods to determine and utilize conceptual relatedness between natural language sources

This invention is a computer program designed to measure how conceptually similar two pieces of text or speech are within a specific context. It converts the language into mathematical representations, then uses calculations to produce a metric of their relatedness. This metric can then be displayed graphically, for example, as a heat map or a score, to show the level of connection. The system can also refine its similarity judgments by comparing the texts against a large knowledge database.

Why it matters: Filed before the significant maturation of large language models (LLMs) and advanced transformer architectures. These technologies now offer highly sophisticated and efficient ways to vectorize natural language and determine semantic relatedness, making the core technical challenge of this invention significantly more achievable and powerful than in 2021.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedSoftware logic, natural language processing, data bank

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