System, device, and method to provide generalized knowledge routing utilizing machine learning to a user within the system
Inventors
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12579368-B2
Publication Date
2026-03-17
Expiration Date
Abstract
The machine learning in the neural networks module can analyze an annotation and its metadata on the annotation made by a first user on a first computing device to make an embedding regarding the annotation and then cooperate with the persistence knowledge store to store the embedding of the machine learning's understanding of the annotation and its metadata. The delivery module can proactively push a notice regarding a potentially related embedding out to a second user on a second computing device based on a threshold amount of relatedness between one or more factors of i) a first task undertaken by the first user and a second task undertaken by the second user, ii) a role of the first user and a role of the second user, and iii) a subject matter of the embedding to a subject matter of a task undertaken by the second user.
Core Innovation
The invention provides generalized knowledge routing using machine learning to users within a system. A delivery module cooperates with a neural networks model and a persistence knowledge store. Machine learning in the neural networks module analyzes an annotation and its metadata made by a first user on a first computing device to make an embedding, and stores the embedding of the machine learning’s understanding of the annotation and its metadata in the persistence knowledge store.
The delivery module facilitates a collaborative process between users by sending a push notice to a second user on a second computing device regarding a potentially related embedding. Whether to send the notice is based on a threshold amount of relatedness between one or more factors selected from a first task and a second task, a role of the first user and a role of the second user, and a subject matter of the embedding to a subject matter of a task undertaken by the second user.
The system further supports generalized knowledge routing evaluation and feedback, where consumption of a proactively delivered notice provides feedback. An evaluation module and a feedback module update stored embedding distances in a shared embedding space, refine the machine learning’s understanding of the embedding, and store the refined understanding back in the persistence knowledge store.
Claims Coverage
The provided material contains three independent claim types: an apparatus claim, a method claim, and a non-transitory machine-readable medium claim. Across the independent claims, the coverage centers on a delivery module cooperating with a neural networks model and a persistence knowledge store to create and store embeddings from an annotation plus metadata, then proactively push a notice to a second user based on a threshold amount of relatedness computed from selected factors.
Delivery module with neural networks model and persistence knowledge store
A delivery module cooperates with a neural networks model and a persistence knowledge store to analyze an annotation and its metadata made by a first user on a first computing device to make an embedding, and store the embedding of the machine learning’s understanding of the annotation and its metadata.
Threshold-based proactive push notice for collaborative knowledge routing
The delivery module facilitates a collaborative process between users by sending a push notice regarding a potentially related embedding to a second user on a second computing device based on a threshold amount of relatedness between a first task and a second task, a role of the first user and a role of the second user, and a subject matter of the embedding to a subject matter of a task undertaken by the second user.
Shared embedding space distance evaluation based on notice consumption
An evaluation module and a feedback module, based on whether a proactively delivered notice is consumed by a second user, update stored embedding distances in a shared embedding space and refine a neural networks model to produce and store a refined understanding of the embeddings in the persistence knowledge store.
Semantic analysis and parsing for file portions to create embeddings
A collections module uses one or more sensors to capture and semantically analyze annotated file content, and supplies a file portion and its annotation to a neural-network machine-learning model to create an embedding.
Synthetic data generated by a generative large language model for embedding-related training
The neural networks model uses synthetic data generated by a generative large language model to create synthetic versions of annotations during an initial training process.
Graph neural network in the neural networks model
The neural network model includes at least one graph neural network.
Query-based exchange between delivery module, neural networks module, and persistence knowledge store
A delivery module exchanges queries with a neural networks module and a persistence knowledge store to retrieve embedding-related information and, based on that determination, proactively push a notice about a potentially related embedding to a second user.
Across the independent claims, the core inventive coverage is the combination of embedding creation from a first user’s annotation and metadata using a neural networks model, storage in a persistence knowledge store, and collaborative proactive notification to a second user using the delivery module. The proactive notice is conditioned on a threshold amount of relatedness computed from tasks, roles, and subject matter, and dependent features further cover evaluation and feedback via shared embedding space distance updates, semantic parsing, synthetic-data generation, inclusion of a graph neural network, and query-based retrieval interactions.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?