System and method for automated course individualization via learning behaviors and natural language processing

• US 10,339,822 B2
• Filed: 07/06/2016
• Issued: 07/02/2019
• Est. Priority Date: 10/26/2012
• Status: Active Grant

• Current Assignee: Zoomi, Inc.
• Original Assignee: Zoomi, Inc.
• Inventors: Brinton, Christopher Greg, Chen, Weiyu, Chiang, Mung, Ha, Sangtae, Rill, Stefan Ruediger
• Primary Examiner(s): Bullington, Robert P
• Application Number:
  US15/202,663
• Publication Number:
  US 20160314699A1
• Time in Patent Office: 1,091 Days
• Field of Search: 434362
• CPC Class Codes: G09B 5/02 with visual presentation of the material to be studied, e.g. using film strip
• CPC Class Codes: G09B 5/12 different stations being capable of presenting different information simultaneously G09B5/14 takes precedence
• CPC Class Codes: G09B 7/00 Electrically-operated teaching apparatus or devices working with questions and answers mechanically operated G09B3/00; computing arrangements G06F

Abstract

A system and method to optimize learning efficacy and efficiency in an online course is disclosed. In particular, the methods include customizing the sequence of delivery of course content as the course is being delivered, in a way that does not necessitate upfront input from an instructor/author or anyone else, beyond what which would be provided for a standard, non-adaptive course already. The present invention is also directed to a system to implement said customization and individualization methods. The present method is further directed to a linear flow of delivered materials, but the flow is dependent upon student actions in the course, among other conditions. In the present invention, individualized adaptation is based on this input, but can be augmented with additional information provided by instructors, if desired, as well.

Claims

1. A method for a multi-engine, processor-based server in communication with a data store and a remote multimedia interactive student graphical user interface (GUI) on a student'"'"'s display, said GUI operating with said display and configured to capture student actions for delivery to said server, to select and deliver interactive course modules to a computer display customized to a student'"'"'s learning abilities determined based on said student'"'"'s actions, said customization occurring during course delivery, comprising the steps of:

organizing the content of a course into modules individually deliverable to a GUI, each said module stored in a data store and including at least one of text, video, and images;

with an engine for processing sound, image, and natural language, identifying the frequency of words and topics in each said module;

with a vector development engine, using said determined frequency of words and topics in each said module, forming a first vector relative to each said module, each said first vector stored in a data store;

with said vector development engine, forming a second vector relative to each module, each said second vector representing individual and sequences of actions, together with times between actions, for prior students interacting with the associated module, together with each prior student'"'"'s determined comprehension of content, and storing each said second vector in a data store, where said actions include at least one of clicks, rewinds, and student generated starts and stops;

with said vector development engine, forming a third vector specific to said student, said third vector representing said student'"'"'s prior actions and times between actions relative to any prior courses taken by said student, together with determined comprehension of said student, and storing said third vector in a data store;

with a delivery engine, delivering a first course module to said student GUI using an internet, for display and interaction on said GUI, said GUI including a programmed capture engine for capturing student actions and time between actions by said student and delivering said captured actions and times between actions to said vector development engine, thereby forming a fourth vector, said fourth vector characterizing said student'"'"'s actions, times between actions, and sequences of actions in said first module,

delivering said first and said second vectors associated with said delivered module and said third and fourth vector to an analysis engine for comparison, said comparison including said captured data analyzed as both individual and sequences of actions and identifying known events associated with comprehension for said student and updating said third vector;

identifying potential next modules to deliver to said student based on an analysis of completed topics and remaining topics in a syllabus;

analyzing said fourth vector to identify sequences of actions known to be associated with a student'"'"'s learning ability;

upon such an event being identified relative to at least one sequence of actions, assessing said student'"'"'s comprehension by topic and type of content of said first module'"'"'s content

further assessing said student'"'"'s learning ability and comprehension by topic and type of content by comparing said fourth vector for the delivered module relative to said first and second vectors relative to the delivered module;

updating said third vector and determining a next module to deliver to said student from said identified potential next modules based on comparison of first and second vectors of potential next modules; and

delivering said determined next module to said student.

2. The method of claim 1, wherein the step of determining is based at least in part on calculating highest likelihood of improvement in percent correct for said student on a first attempt in quizzes in the next module.

3. The method of claim 1, wherein each said sequences of actions is characterized as including combinations of any one of reflecting, reviewing, skimming, or speeding and said best fit comparison includes an analysis of said sequences against similar sequences associated with known learning abilities.

4. The method of claim 1, wherein in the step of determining, extended intervals are discounted.

5. The method of claim 1, wherein said comparison is based on cosine similarity.

6. The method of claim 1, wherein text identification includes natural language processing.

7. The method of claim 1, wherein said GUI further captures biosensed bodily movements together with time between events and said bodily movements are used for formulating said fourth vector of said student'"'"'s actions.

8. The method of claim 7, wherein student bodily movements include eye and hand movements, and events include said student changing status between in front of and away from the display.

9. The method of claim 8, wherein said student'"'"'s eye and body movements are captured by a camera associated with said computer display.

10. The method of claim 9, wherein said assessing includes an analysis of said student'"'"'s eye and body movements.

11. A method for a server with access to a data store and an interactive GUI player to deliver course content customized to a student for presentation on a display using the internet comprising the steps of:

in a stored collection of course modules, identifying and storing by text term the frequency of text terms in each module in said collection of modules;

delivering a first course module to a player in control of said student for display and interaction;

formulating a module vector for each course module, each module vector arranged algebraically by text term describing course content of its associated module;

tracking student behaviors relative to said first course module using said player and said server receiving said student behaviors from said player;

formulating a student vector, arranged by text term, describing the student'"'"'s tracked behaviors, said student'"'"'s tracked behaviors including clicks, rollovers, time between events, quiz responses, and eye and body movements and said student vector being characterized by frequency of text terms in said student'"'"'s tracked behaviors;

selecting a next module for delivery based on a statistical comparison of each of said module vectors to said student vector and a comparison of completed topics to a course syllabus; and

delivering said next module to said student.

12. The method of claim 11, wherein each said module vector is further calculated using an automated thesaurus so as to aggregate like-meaning terms to determine frequency of terms for said vector.

13. The method of claim 11, wherein each said module vector includes weightings based on words appearing in a course syllabus.

14. The method of claim 11, wherein said student vector has weighting factors based on frequency of student rewinds.

15. The method of claim 11, wherein said student vector includes an accounting for student tracked behaviors in previously viewed modules by said student.

16. The method of claim 11 wherein each said module vector is further formulated based on converting any video or audio content to text.

17. A system for selecting and sequencing delivery of course modules over an internet, customized to a student, comprising;

a processor-based server;

a data store; and

a display for displaying course content for student interaction and for collecting and forwarding student behavior events together with time between student behavior events, where said student behavior events include clicks, rollovers, and video control;

wherein said data store is used for storing said course content in modular form; and

said processor-based server is used to convert audio and video to text, store said text and the frequency of said text with the related module, calculate a vector representing text content of each module, use tracked student behavior events for formulating a student vector, calculate a student vector in consistency with said behavior events, and perform a statistical analysis of said student vector relative to each module vector to determine which next module to deliver to said student.

18. The system of claim 17, wherein said display further includes recording means for visually recording physical movements of said student and said server includes said physical movements in tracked student behaviors.

19. The system of claim 17, wherein said display further includes recording means for visually recording eye movements of said student and said server includes said eye movements in said tracked student behaviors.

20. The system of claim 17, wherein said student vector has weighting factors based on frequency of student rewinds.