Claims:WE CLAIM
1. A method of generating a learning path for a topic, the method comprising:
extracting, by a learning path generation device, a plurality of key phrases from each of a plurality of learning resources related to the topic;
determining, by the learning path generation device, a learning context for each of the plurality of learning resources based on the plurality of key phrases;
forming, by the learning path generation device, a set of key phrase groups from among the plurality of key phrases for each of the plurality of learning resources;
determining, by the learning path generation device, a relationship among the key phrases in each of the set of key phrase groups based on the learning context;
generating, by the learning path generation device, a structured graph for the plurality of learning resources based on the plurality of key phrases and the relationship among the key phrases; and
generating, by the learning path generation device, the learning path for the topic based on the structured graph for the plurality of learning resources.

2. The method of claim 1, further comprising:
acquiring the plurality of learning resources related to the topic; and
storing the plurality of learning resources in a database.

3. The method of claim 1, wherein the plurality of key phrases is extracted, using a first machine learning algorithm, by:
determining a set of n-gram frequencies based on a set of tags in a given learning resource, and
determining the plurality of key phrases based on the set of n-gram frequencies.

4. The method of claim 1, wherein the learning context is determined using a second machine learning algorithm, and wherein the second machine learning algorithm is a long short-term memory (LSTM) machine learning algorithm.

5. The method of claim 1, wherein key phrase groups comprises key phrase pairs, and wherein each of the set of key phrase pairs is formed by pairing each of the plurality of key phrases of a given learning resource with at least one of: each of a plurality of remaining key phrases of the given learning resource, or each of a plurality of randomly selected key phrases from a corpus of key phrases related to the topic.

6. The method of claim 1, wherein the relationship among the key phrases in a given key phrase group is determined, using a third machine learning algorithm, by:
determining a probability of occurrence of a key phrase in the given key phase group in presence of remaining key phrases in the given key phase group; and
determining a nature of relationships among the key phrases, in the given key phase group, in a multi-dimension space.

7. The method of claim 6, further comprising:
validating, using a fourth and a fifth machine learning algorithm, the relationship among the key phrases in the given key phrase group; and
tuning the third machine learning algorithm based on the validation.

8. The method of claim 6, wherein the nature of relationship comprises a parent-child relationship, a prerequisite-post requisite relationship, a related relationship, or an independent relationship.

9. The method of claim 1, further comprising generating a personalized learning path for a user based on a current understanding of the user on the topic and the learning path for the topic, wherein the current understanding of the user is determined based on an input from the user on one or more prerequisites identified for the topic.

10. The method of claim 9, further comprising determining a learning time for the user based on the personalized learning path.

11. A system for generating a learning path for a topic, the system comprising:
a learning path generation device comprising at least one processor and a computer-readable medium storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations comprising:
extracting a plurality of key phrases from each of a plurality of learning resources related to the topic;
determining a learning context for each of the plurality of learning resources based on the plurality of key phrases;
forming a set of key phrase groups from among the plurality of key phrases for each of the plurality of learning resources;
determining a relationship among the key phrases in each of the set of key phrase groups based on the learning context;
generating a structured graph for the plurality of learning resources based on the plurality of key phrases and the relationship among the key phrases; and
generating the learning path for the topic based on the structured graph for the plurality of learning resources.

12. The system of claim 11, wherein the plurality of key phrases is extracted, using a first machine learning algorithm, by:
determining a set of n-gram frequencies based on a set of tags in a given learning resource, and
determining the plurality of key phrases based on the set of n-gram frequencies.

13. The system of claim 11, wherein the learning context is determined using a second machine learning algorithm, and wherein the second machine learning algorithm is a long short-term memory (LSTM) machine learning algorithm.

14. The system of claim 11, wherein key phrase groups comprises key phrase pairs, and wherein each of the set of key phrase pairs is formed by pairing each of the plurality of key phrases of a given learning resource with at least one of: each of a plurality of remaining key phrases of the given learning resource, or each of a plurality of randomly selected key phrases from a corpus of key phrases related to the topic.

15. The system of claim 11, wherein the relationship among the key phrases in a given key phrase group is determined, using a third machine learning algorithm, by:
determining a probability of occurrence of a key phrase in the given key phase group in presence of remaining key phrases in the given key phase group; and
determining a nature of relationships among the key phrases, in the given key phase group, in a multi-dimension space.

16. The system of claim 15, wherein the operations further comprise:
validating, using a fourth and a fifth machine learning algorithm, the relationship among the key phrases in the given key phrase group; and
tuning the third machine learning algorithm based on the validation.

17. The system of claim 15, wherein the nature of relationship comprises a parent-child relationship, a prerequisite-postrequisite relationship, a related relationship, or an independent relationship.

18. The system of claim 11, wherein the operations further comprise generating a personalized learning path for a user based on a current understanding of the user on the topic and the learning path for the topic, and wherein the current understanding of the user is determined based on an input from the user on one or more prerequisites identified for the topic.

19. The system of claim 18, wherein the operations further comprise determining a learning time for the user based on the personalized learning path.

Dated this 10th day of October, 2018

R Ramya Rao
Of K&S Partners
Agent for the Applicant
IN/PA-1607
, Description:TECHNICAL FIELD
This disclosure relates generally to information processing, and more particularly to method and system for generating a learning path using machine learning.