System and method for hybrid question answering over knowledge graph

This invention describes a computer system that answers questions asked in everyday language. It works by identifying the main topic of a question within a large network of interconnected facts, called a knowledge graph. The system then uses a specific type of artificial intelligence model, a BERT encoder, to deeply understand both the question and related information from the graph, ultimately selecting the most relevant answer. The claims specifically detail the use and training of this BERT model.

Why it matters: Filed before the widespread adoption of large language models (LLMs), the specialized BERT encoding and training described here can now be achieved with greater power and often less custom effort using more advanced, general-purpose LLMs. This makes the core functionality of question answering over a knowledge graph more accessible and effective.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedAI/ML software development, knowledge graph

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