Skill shortlister for natural language processing

This invention describes a system that takes a spoken command, converts it into text, and also creates a numerical representation of the audio. It then uses multiple specialized decision-makers to calculate the probability that the command corresponds to different specific applications or "skills," like playing music or setting a timer. Based on these probabilities, it selects the most likely skill. The claims specifically describe a method for processing user-spoken commands, incorporating user-specific skill enablement data, and training the system to improve its accuracy.

Why it matters: Filed before the widespread adoption of advanced transformer models and large language models. These technologies have dramatically improved the accuracy and flexibility of generating vector representations and training classifiers for natural language understanding, making the core task of skill shortlisting more robust and scalable than in 2018.

Status: Not in forceListed as no longer active.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

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