System, device, and method to provide generalized knowledge routing utilizing machine learning to a user within the system
View Patent ↗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.
1 . An apparatus, comprising:
a delivery module that is configured to cooperate with a neural networks model, and a persistence knowledge store,
where machine learning in the neural networks module is configured to 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, and
where the delivery module is further configured to facilitate a collaborative process between users of a system 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 one or more factors selected from a group consisting 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, where portions of the delivery module, the neural networks model, and the persistence knowledge store implemented in software instructions are stored in an executable format in one or more memories to be executed by one or more processing units.
2 . The apparatus of claim 1 , further comprising:
a collections module is configured to use one or more sensors to capture and understand content in the annotation and portions of a file corresponding to the annotation, and then after a semantic analysis and parsing of the content in the portion of the file and its corresponding annotation, then the collections module is configured to supply the portion of the file and its corresponding annotation for the analysis using the machine learning in the neural networks model to create the embedding.
3 . The apparatus of claim 1 , where the delivery module is configured to cooperate with at least one of the persistence knowledge store and the neural networks model to monitor activities of the second user in order to trigger a proactive push of the notice regarding the potentially related embedding.
4 . The apparatus of claim 1 , where the neural networks model is configured to use synthetic data coming from a generative large language model to generate synthetic versions of the annotations with the synthetic data in an initial training process of the neural networks model.
5 . The apparatus of claim 1 , where the delivery module is further configured to cooperate with at least one of the persistence knowledge store and the neural networks model, where the delivery module is configured to make queries back and forth with at least one of the neural networks module and the persistence knowledge store in order to retrieve information regarding the embedding on the annotation its meta data and a portion of a file corresponding to the annotation, when making a determination to proactively push the notice regarding the potentially related embedding out to the second user on the second computing device.
6 . The apparatus of claim 1 , further comprising:
an evaluation module configured to track whether the second user consumed the proactively delivered notice.
7 . The apparatus of claim 6 , where the evaluation module is configured to cooperate with a feedback module such that the consumption event of the proactively delivered notice is then supplied by the feedback module to at least one of i) the persistence knowledge store to adjust a distance between stored embeddings in a common shared embedding space relative to other embeddings based upon whether the proactively delivered notice was consumed by the second user or not, and ii) to the neural networks model to refine the machine learning's understanding of the embedding with additional information about the embedding based upon whether the proactively delivered notice was consumed by the second user or not and then store the refined machine learning's understanding of the embedding in the common shared embedding space relative to other embeddings in the persistence knowledge store.
8 . The apparatus of claim 1 , where the collection module and the neural networks model are configured to cooperate with the persistence knowledge store to create a shared embedding space for the embedding relative to other embeddings, as well as the persistence knowledge store has a search function for all of the embeddings stored in that persistence knowledge store, where a corpus of all of the embeddings were produced by the neural networks model.
9 . The apparatus of claim 1 , where the delivery module is further configured to proactively push the notice regarding the potentially related embedding out to the second user on the second computing device based on the threshold amount of relatedness between all three factors selected from the group consisting of i) the first task undertaken by the first user and the second task undertaken by the second user, ii) the role of the first user and the role of the second user, and iii) the subject matter of the embedding to the subject matter of task undertaken by the second user.
10 . The apparatus of claim 1 , where the neural networks model is composed of at least one or more graph neural networks.
11 . A method to provide generalized knowledge routing utilizing machine learning to users within a system, comprising:
providing a delivery module that cooperates with a neural networks model, and a persistence knowledge store,
providing machine learning in the neural networks module to 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, and
providing the delivery module to proactively push a notice regarding a potentially related embedding to a second user on a second computing device based on a threshold amount of relatedness between one or more factors selected from a group consisting 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, where portions of the delivery module, the neural networks model, and the persistence knowledge store implemented in software instructions are stored in an executable format in one or more memories to be executed by one or more processing units.
12 . The method of claim 11 , further comprising:
providing a collections module to use one or more sensors to capture and understand content in the annotation and portions of a file corresponding to the annotation, and then after a semantic analysis and parsing of the content in the portion of the file and its corresponding annotation, then the collections module to supply the portion of the file and its corresponding annotation for the analysis using the machine learning in the neural networks model to create the embedding.
13 . The method of claim 11 , further comprising:
providing the delivery module to cooperate with at least one of the persistence knowledge store and the neural networks model to monitor activities of the second user in order to proactively push the notice regarding the potentially related embedding.
14 . The method of claim 11 , further comprising:
providing the neural networks model to use synthetic data coming from a generative large language model to generate synthetic versions of the annotations with the synthetic data in an initial training process of the neural networks model.
15 . The method of claim 11 , further comprising:
providing the delivery module to cooperate with at least one of the persistence knowledge store and the neural networks model, where the delivery module is configured to make queries back and forth with at least one of the neural networks module and the persistence knowledge store in order to retrieve information regarding the embedding on the annotation its meta data and a portion of a file corresponding to the annotation, when making a determination to proactively push the notice regarding the potentially related embedding out to the second user on the second computing device.
16 . The method of claim 11 , further comprising:
providing an evaluation module to track whether the second user finds the proactively delivered notice on the annotation useful by determining whether the second user consumed the proactively delivered notice.
17 . The method of claim 16 , further comprising:
providing the evaluation module to cooperate with a feedback module such that the consumption event of the proactively delivered notice is then supplied by the feedback module to at least one of i) the persistence knowledge store to adjust a distance between stored embeddings in a common shared embedding space relative to other embeddings based upon whether the proactively delivered notice was consumed by the second user or not, and ii) to the neural networks model to refine the machine learning's understanding of the embedding with additional information about the embedding based upon whether the proactively delivered notice was consumed by the second user or not and then store the refined machine learning's understanding of the embedding in the common shared embedding space relative to other embeddings in the persistence knowledge store.
18 . The method of claim 11 , further comprising:
providing the delivery module and the neural networks model to cooperate with the persistence knowledge store to create a shared embedding space for the embedding relative to other embeddings, as well as the persistence knowledge store has a search function for all of the embeddings stored in that persistence knowledge store, where a corpus of all of the embeddings were produced by the neural networks model.
19 . The method of claim 11 , further comprising:
providing the delivery module to proactively push the notice regarding the potentially related embedding out to the second user on the second computing device based on the threshold amount of relatedness between all three factors selected from the group consisting of i) the first task undertaken by the first user and the second task undertaken by the second user, ii) the role of the first user and the role of the second user, and iii) the subject matter of the embedding to the subject matter of task undertaken by the second user.
20 . A non-transitory machine-readable medium, which stores further instructions in an executable format by one or more processors to cause operations as follows, comprising:
using a delivery module that cooperates with a neural networks model, and a persistence knowledge store,
using machine learning in the neural networks module to 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, and
using the delivery module to 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 selected from a group consisting 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, where portions of the delivery module, the neural networks model, and the persistence knowledge store implemented in software instructions are stored in an executable format in one or more memories to be executed by one or more processing units.