Intelligent tutoring system

• US 8,684,747 B1
• Filed: 12/14/2012
• Issued: 04/01/2014
• Est. Priority Date: 12/10/2007
• Status: Active Grant

• Current Assignee: Accella Learning, LLC
• Original Assignee: Accella Learning, LLC
• Inventors: Bergeron, Bryan
• Primary Examiner(s): EGLOFF, PETER RICHARD
• Application Number:
  US13/714,474
• Time in Patent Office: 473 Days
• Field of Search: 434/323, 434/350, 434/362, 434/365, 707/602, 707/625
• US Class Current: 434/323
• CPC Class Codes: G09B 7/04 characterised by modifying the teaching programme in response to a wrong answer, e.g. repeating the question, supplying a further explanation

Abstract

An intelligent tutoring system can be implemented using filters, predictive modeling, and a knowledge warehouse, which allows content to be dynamically selected for individualized presentation to a learner. Such an intelligent tutoring system can be implemented using a variety of techniques, including polynomial vector equations in modeling, extensible markup language storage formats for presentation data, and a knowledge warehouse with information stored in data cubes.

Claims

1. A non-transitory computer readable medium having stored thereon a set of computer executable instructions operable to configure a computer to perform a set of acts comprising:

a) receiving a set of user data from a database, said user data comprising a plurality of values organized according to a plurality of competency areas;

b) storing said set of received user data in a local memory of said computer;

c) applying the set of received user data stored in said local memory to a predictive model;

d) determining a set of content for presentation to a user, wherein;

i) the set of content comprises educational material selected to maximize a result indicated by the predictive model, wherein maximizing a result indicated by said predictive model comprises matching material for presentation to the user with a predicted skill level for the user at the time of presentation as indicated by the predictive model; and

ii) the set of content is tagged according to a first competency area from the plurality of competency areas;

e) determining an evaluation for presentation to said user, wherein said evaluation comprises at least one question tagged according to the first competency area from the plurality of competency areas;

f) incorporating a set of evaluation data received from said user in response to said evaluation into said set of user data, wherein incorporating said set of evaluation data comprises;

i) normalizing said set of evaluation data by performing a cleaning operation and a transformation operation on the set of evaluation data, wherein both the cleaning operation and the transformation operation are specified in a metadata scheme associated with the user data in the database; and

ii) storing said normalized set of evaluation data with said set of received user data.

2. The computer readable medium of claim 1 wherein said predictive model comprises a nonlinear polynomial equation.

3. The computer readable medium of claim 1, wherein performing the transformation operation on the set of evaluation data comprises modifying an order for answers in the set of evaluation data.

4. The computer readable medium of claim 1, wherein performing the transformation operation on the set of evaluation data comprises performing a conversion from a first format to a second format.

5. The computer readable medium of claim 4, wherein performing the conversion from the first format to the second format comprises converting a date of birth for the user from the first format to the second format.

6. A method of automatically selecting educational content to present to a student and storing information about the student'"'"'s educational achievement comprising:

a) using a set of computer executable instructions stored in local memory on a computer, causing the computer to execute a retrieval module to retrieve a competency record for the student from a database, wherein said competency record comprises a set of data indicating said student'"'"'s knowledge in a plurality of competencies, and wherein retrieving the competency record for the student comprises storing said set of data in the computer'"'"'s local memory;

b) using the set of computer executable instructions, causing the computer to execute a transition module to determine a set of educational content for said student based on a comparison of a predicted skill level for the student indicated by a first predictive model comprising a nonlinear polynomial student model and a level for said set of educational content in a first competency from said plurality of competencies, wherein said level is retrieved from metadata associated with said set of educational content;

c) using the set of computer executable instructions, causing the computer to execute an evaluation module to dynamically select a plurality of questions assessing knowledge in said first competency for presentation to said student;

d) using the set of computer executable instructions, causing the computer to execute a normalization module to perform a cleaning operation and a transformation operation on a set of evaluation data, wherein said set of evaluation data is based on input received from said student in response to the presentation of the plurality of questions; and

e) using the set of computer executable instructions, causing the computer to execute an update module to modify the set of data stored in the computer'"'"'s local memory to reflect set of evaluation data and the cleaning and transformation operations performed by the normalization module and to propagate the modification to the set of data back to the database;

wherein;

A) the set of evaluation data is used, during presentation of the set of educational content to the student, to update the first predictive model; and

B) the method comprises, during presentation of the set of educational content, the updated first predictive model indicating a second predicted skill level for the student.

7. The method of claim 6 wherein the first predictive model comprises a polynomial vector student model.

8. The method of claim 6, wherein:

a) performing said cleaning operation comprises applying a rule for allowable values to said set of evaluation data;

b) performing said transformation operation comprises converting a test result from said set of evaluation data into a z-score; and

c) the cleaning operation and the transformation operation are indicated in the metadata associated with the set of educational content.