Personalized learning system

1. A learning system comprising:

a non-transitory memory; and

one or more hardware processors configured or programmed to read instructions from the non-transitory memory to cause the learning system to perform operations including:

generating a user knowledge mesh including:

generating a plurality of topic nodes, each of the plurality of topic nodes corresponding to a topic included in the user knowledge mesh; and

generating a plurality of concept nodes, each of the plurality of concept nodes corresponding to a key learnable concept;

updating the user knowledge mesh by deleting all existing connections between the plurality of topic nodes; and

generating new connections between the plurality of topic nodes based on the semantic similarities between the topics corresponding to the plurality of topic nodes; wherein

each of the plurality of topic nodes is connected to another one of the plurality of topic nodes;

each of the plurality of concept nodes is connected to one of the plurality of topic nodes;

each of the plurality of key learnable concepts includes one or more interactions related to the key learnable concept; and

the deleting of all the existing connections between the plurality of topic nodes and the generating of the new connections between the plurality of topic nodes are performed when a predetermined number of new topics have been added to the user knowledge mesh.

2. The learning system of claim 1 , wherein the operations further include:

calculating a semantic similarity between a first topic that corresponds to a first topic node of the plurality of topic nodes and a second topic that corresponds to a second topic node of the plurality of topic nodes; and

generating a line that connects the first topic node and the second topic node; wherein

a length, a thickness, and/or a brightness of the line is generated based on the semantic similarity between the first topic and the second topic.

3. The learning system of claim 1 , wherein the operations further include:

changing a color of one of the plurality of concept nodes when a user readiness for the key learnable concept that corresponds to the one of the plurality of concept nodes is above a predetermined readiness threshold.

4. The learning system of claim 1 , wherein the operations further include:

generating a new topic to include in the user knowledge mesh including:

identifying a desired topic that is input by a user of the user knowledge mesh and that corresponds to the new topic;

determining whether or not the desired topic input by the user already exists in the learning system by determining if a previously created topic that corresponds to the desired topic is already stored in the non-transitory memory; and

adding the previously created topic that corresponds to the desired topic to the user knowledge mesh when it is determined that the desired topic input by the user already exists in the learning system; wherein

the previously created topic that corresponds to the desired topic is determined to already be stored in the non-transitory memory based on the previously created topic being a lexical match to the desired topic and/or a vector distance between a representation of the previously created topic and a representation of the desired topic in a semantic embedding model.

5. The learning system of claim 4 , wherein the determining whether or not the desired topic input by the user already exists in the learning system includes measuring a relevancy of each of a plurality of previously created and stored topics to content already included in the user knowledge mesh.

6. The learning system of claim 1 , wherein the operations further include:

generating a new topic to include in the user knowledge mesh including:

identifying a desired topic that is selected by a user of the user knowledge mesh and that corresponds to the new topic; wherein

the desired topic is selected directly from a list of topics that includes a plurality of topics that have already been created and stored in the non-transitory memory; and

the plurality of topics included in the list of topics have been filtered based on an average or minimum quality.

7. The learning system of claim 1 , wherein the operations further include:

generating a new topic to include in the user knowledge mesh including:

identifying a desired topic that is input by a user of the user knowledge mesh and that corresponds to the new topic;

searching for the desired topic in an open source data set to identify a related open source document;

using natural language processing to extract one or more new key learnable concepts from the related open source document; and

generating one or more new interactions for each of the one or more new key learnable concepts.

8. The learning system of claim 7 , wherein the generating the one or more new interactions includes:

using natural language processing and the related open source document to generate a definition for each of the one or more new key learnable concepts;

generating, for each of the one or more new key learnable concepts, one or more distractors based on the definition of the new key learnable concept; and

generating, for each of the one or more new key learnable concepts, the one or more new interactions based on the definition of the new key learnable concept and the one or more distractors.

9. The learning system of claim 1 , wherein the operations further include:

generating a new topic and a corresponding new topic node;

calculating a sematic similarity between the new topic and each of the plurality of topics that already exist in the user knowledge mesh; and

connecting the new topic node to one of the plurality of topic nodes corresponding to the one of the plurality of topics with a highest semantic similarity to the new topic.

10. The learning system of claim 1 , wherein the operations further include:

generating a new topic and a corresponding new topic node;

determining a number of connections of the new topic node and a number of connections of one of the plurality of topic nodes that already exists in the user knowledge mesh; and

determining whether or not to connect the new topic node to the one of the plurality of topic nodes that that already exists in the user knowledge mesh based on a sum of the number of connections of the new topic node and the number of connections of the one of the plurality of topic nodes that already exists in the user knowledge mesh.

11. The learning system of claim 1 , wherein the operations further include:

generating a new topic based on the key learnable concepts that already exist in the user knowledge mesh; wherein

the generating the new topic includes:

identifying a desired topic input by a user of the user knowledge mesh;

identifying one or more topics that already exist in the user knowledge mesh and are related to the desired topic based on semantic similarities between the one or more topics that already exist in the user knowledge mesh and the desired topic;

identifying one or more key learnable concepts from the one or more topics identified based on semantic similarities between the one or more key learnable concepts and the desired topic; and

creating the new topic using a predetermined number of the one or more key learnable concepts identified.

12. The learning system of claim 1 , wherein the operations further include:

generating a new topic based on the plurality of topics that already exist in the user knowledge mesh; wherein

the generating the new topic includes:

calculating a semantic similarity score between each of a plurality of potential new topics and one or more goals of a user of the user knowledge mesh;

calculating a semantic similarity score between each of the plurality of potential new topics and the plurality of topics that already exist in the user knowledge mesh;

calculating, for each of the plurality of potential new topics, a weighted sum in which the semantic similarity between the potential new topic and the one or more goals of the user is added and the semantic similarity between the potential new topic and the plurality of topics that already exist in the user knowledge mesh is subtracted; and

selecting the potential new topic that has a highest weighted sum as the new topic.

13. The learning system of claim 12 , wherein the operations further include:

calculating a quality of the plurality of topics that already exist in the user knowledge mesh; and

generating the new topic based on the quality of the plurality of topics that already exist in the user knowledge mesh; wherein

the quality of the plurality of topics that already exist in the user knowledge mesh are calculated based on a quality of the one or more interactions included in the plurality of topics.

14. The learning system of claim 1 , wherein the operations further include:

generating a key learnable concept that corresponds to one of the plurality of concept nodes using a document and/or a file identified by a user of the learning system; wherein

the generating the key learnable concept includes:

identifying content of the document and/or the file; and

identifying the key learnable concept from the content of the document and/or the file.

15. The learning system of claim 1 , wherein the operations further include:

generating a key learnable concept that corresponds to one of the plurality of concept nodes using at least one of a videoconferencing platform, a teleconferencing platform, and an application; wherein

the generating the key learnable concept includes:

identifying content of a meeting based on an input received by a user device; and

identifying the key learnable concept from the content of the meeting.

16. The learning system of claim 15 , wherein

the input received by the user device identifies a particular point in time during the meeting; and

a predetermined amount of content of the meeting from prior to the input being received by the user device is identified as the content of the meeting from which the key learnable concept is identified.

17. The learning system of claim 1 , wherein the operations further include:

determining a readiness score of a user with respect to each of the plurality of key learnable concepts; and

automatically navigating the user through the user knowledge mesh by directing the user to the concept node that corresponds to the key learnable concept for which the user has a lowest readiness score; wherein

the determining the readiness score of the user with respect to each of the plurality of key learnable concept includes:

calculating an age value corresponding to an amount of time since the key learnable concept was last reviewed by the user;

calculating an interim knowledge state value based on the age value and a half-life value that represents an estimated half-life of the memory of the user with respect to the key learnable concept; and

calculating the readiness score of the user with respect to the key learnable concept based on the interim knowledge state value and a number of times the user has reviewed the key learnable concept.

18. The learning system of claim 1 , wherein the operations further include:

determining a semantic similarity between each of the plurality of key learnable concepts and one or more goals of the user; and

automatically navigating the user through the user knowledge mesh by directing the user to the concept node that corresponds to the key learnable concept that has a highest semantic similarity to the one or more goals of the user.

19. A learning system comprising:

a non-transitory memory; and

one or more hardware processors configured or programmed to read instructions from the non-transitory memory to cause the learning system to perform operations including:

generating a user knowledge mesh including:

generating a plurality of topic nodes, each of the plurality of topic nodes corresponding to a topic included in the user knowledge mesh; and

generating a plurality of concept nodes, each of the plurality of concept nodes corresponding to a key learnable concept; and

generating a new topic based on the plurality of topics that already exist in the user knowledge mesh; wherein

each of the plurality of topic nodes is connected to another one of the plurality of topic nodes;

each of the plurality of concept nodes is connected to one of the plurality of topic nodes; and

each of the plurality of key learnable concepts includes one or more interactions related to the key learnable concept; and

the generating the new topic includes:

calculating a semantic similarity score between each of a plurality of potential new topics and one or more goals of a user of the user knowledge mesh;

calculating a semantic similarity score between each of the plurality of potential new topics and the plurality of topics that already exist in the user knowledge mesh;

calculating, for each of the plurality of potential new topics, a weighted sum in which the semantic similarity between the potential new topic and the one or more goals of the user is added and the semantic similarity between the potential new topic and the plurality of topics that already exist in the user knowledge mesh is subtracted;

selecting the potential new topic that has a highest weighted sum as the new topic;

adjusting a weight of the semantic similarity score between each of the plurality of potential new topics and the one or more goals of the user to control how closely the selected new topic is to the one or more goals of the user; and

adjusting a weight of the semantic similarity score between each of the plurality of potential new topics and the plurality of topics that already exist in the user knowledge mesh to control how similar the selected new topic is to the plurality of topics that already exist in the user knowledge mesh.

20. A learning system comprising:

a non-transitory memory; and

one or more hardware processors configured or programmed to read instructions from the non-transitory memory to cause the learning system to perform operations including:

generating a user knowledge mesh including:

generating a plurality of topic nodes, each of the plurality of topic nodes corresponding to a topic included in the user knowledge mesh; and

generating a plurality of concept nodes, each of the plurality of concept nodes corresponding to a key learnable concept;

determining a readiness score of a user with respect to each of the plurality of key learnable concepts;

determining a semantic similarity between each of the plurality of key learnable concepts and one or more goals of the user;

calculating, for each of the plurality of key learnable concepts, a weighted sum in which the semantic similarity score between the key learnable concept and the one or more goals of the user has a positive weight and the readiness score of the user with respect to the key learnable concept has a negative weight; and

automatically navigating the user through the user knowledge mesh by directing the user to the concept node that corresponds to the key learnable concept with a highest weighted sum; wherein

each of the plurality of topic nodes is connected to another one of the plurality of topic nodes;

each of the plurality of concept nodes is connected to one of the plurality of topic nodes; and

each of the plurality of key learnable concepts includes one or more interactions related to the key learnable concept;

the determining the readiness score of the user with respect to each of the plurality of key learnable concept includes:

calculating an age value corresponding to an amount of time since the key learnable concept was last reviewed by the user;

calculating an interim knowledge state value based on the age value and a half-life value that represents an estimated half-life of the memory of the user with respect to the key learnable concept; and

calculating the readiness score of the user with respect to the key learnable concept based on the interim knowledge state value and a number of times the user has reviewed the key learnable concept.

21. A learning system comprising:

a non-transitory memory; and

one or more hardware processors configured or programmed to read instructions from the non-transitory memory to cause the learning system to perform operations including:

determining a readiness score of a user with respect to a key learnable concept;

when a new key learnable concept has not been reviewed by the user, determining a predetermined number of key learnable concepts that have been reviewed by the user and are most semantically similar to the new key learnable concept that has not been reviewed by the user; and

averaging a plurality of readiness scores of the user with respect to the predetermined number of key learnable concepts that have been reviewed by the user to calculate a readiness score of the user with respect to the new key learnable concept that has not been reviewed by the user; wherein

the determining the readiness score of the user with respect to the key learnable concept includes:

selecting a specific moment in time;

calculating an age value based on the specific moment in time selected, the age value corresponding to an amount of time since the key learnable concept was last reviewed by the user;

calculating an interim knowledge state value based on the age value and a half-life value that represents an estimated half-life of the memory of the user with respect to the key learnable concept; and

calculating the readiness score of the user with respect to the key learnable concept based on the interim knowledge state value and a number of times the user has reviewed the key learnable concept.

22. The learning system of claim 21 , wherein the operations further include:

determining the readiness score of the user with respect to the key learnable concept when the user responds to an interaction related to the key learnable concept, the readiness score corresponding to a predicted accuracy of the interaction;

updating the number of times the user has reviewed the key learnable concept;

determining an actual accuracy of the interaction; and

updating the half-life value based on the actual accuracy of the interaction.

23. The learning system of claim 21 , wherein the readiness scores of the user with respect to the predetermined number of the key learnable concepts that have been reviewed by the user are weighted according to their semantic similarity to the new key learnable concept when the readiness score of the user with respect to the new key learnable concept is calculated.

24. The learning system of claim 21 , wherein the readiness score of the user is based on one or more of a modifier offset based on an interaction difficulty, a modifier offset based on a concept difficulty, and a modifier offset based on a user agility of the user.

25. The learning system of claim 24 , wherein the operations further include:

calculating an error value based on the actual accuracy of the interaction and the predicted accuracy of the interaction; and

updating the one or more of the modifier offset based on the interaction difficulty, the modifier offset based on the concept difficulty, and the modifier offset based on the user agility based on the error value.

26. The learning system of claim 21 , wherein

the readiness score of the user with respect to the key learnable concept is determined for a specific moment in time in the future;

the readiness score is calculated based on an assumption that the user will have an optimal ongoing pattern of interactions with the key learnable concept or an assumption that the user will have a pattern of interactions with the key learnable concept that is consistent with a user's previous pattern of interactions with the key learnable concept; and

when the readiness score is calculated, the age value, the half-life value, and the number of times the user has reviewed the key learnable concept are determined based on simulating the optimal ongoing pattern of interactions or based on simulating the pattern of interactions with the key learnable concept that is consistent with the user's previous pattern of interactions with the key learnable concept.

27. The learning system of claim 21 , wherein the operations further include:

calculating a plurality of readiness scores of the user with respect to a plurality of key learnable concepts; and

calculating a readiness score of the user with respect to a topic defined by the plurality of key learnable concepts by averaging the plurality of readiness scores of the user with respect to the plurality of key learnable concepts.

28. A learning system comprising:

a non-transitory memory; and

one or more hardware processors configured or programmed to read instructions from the non-transitory memory to cause the learning system to perform operations including:

determining, for each of a plurality of users, a plurality of readiness scores with respect to a plurality of key learnable concepts;

calculating, for each of the plurality of users, a readiness score for a topic defined by the plurality of key learnable concepts by averaging the plurality of readiness scores with respect to the plurality of key learnable concepts;

identifying a predetermined number of the plurality of users with the highest readiness scores for the topic defined by the plurality of key learnable concepts; wherein

the determining the plurality of readiness scores with respect to the plurality of key learnable concepts includes:

selecting a specific moment in time;

calculating an age value based on the specific moment in time selected, the age value corresponding to an amount of time since the key learnable concept was last reviewed by the user;

calculating an interim knowledge state value based on the age value and a half-life value that represents an estimated half-life of the memory of the user with respect to the key learnable concept; and

calculating the readiness score of the user with respect to the key learnable concept based on the interim knowledge state value and a number of times the user has reviewed the key learnable concept.