System and method for recommending educational resources

• US 8,699,939 B2
• Filed: 12/19/2008
• Issued: 04/15/2014
• Est. Priority Date: 12/19/2008
• Status: Expired due to Fees

• Current Assignee: PARC-Palo Alto Research Center, Palo Alto Research Center, Inc. (Xerox Holdings Corp.), Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Palo Alto Research Center, Inc. (Xerox Holdings Corp.), Xerox Corporation (Xerox Holdings Corp.)
• Inventors: German, Kristine A., Lofthus, Robert M., Price, Robert Roy, Perronnin, Florent, Cifarelli, Claudio, Do, Minh, Zhou, Rong
• Primary Examiner(s): GEBREMICHAEL, BRUK A
• Application Number:
  US12/339,979
• Publication Number:
  US 20100159432A1
• Time in Patent Office: 1,943 Days
• Field of Search: 434/322, 434/323, 434/350, 434/362
• US Class Current: 434/350
• CPC Class Codes: G06Q 10/10 Office automation; Time management

Abstract

An educational recommender system and a method are provided. The method includes receiving a request to recommend a course of action related to a plurality of current students; accessing a computer database storing student data that corresponds to the plurality of current students; clustering in a computer process the plurality of current students into at least two clusters based at least on granular assessment data associated with student data corresponding to respective current students; and outputting the results of the clustering to a user. The granular assessment data includes a result of an assessment administered to respective students of the plurality of current students, and each assessment includes a plurality of questions for assessing one of the current students. The associated result includes an independent evaluation of each respective question of the plurality of questions.

Claims

1. A recommender system for recommending clustering of students into educational groups, the recommender system comprising:

a processor for executing a series of programmable instructions for;

receiving a request to recommend a course of action related to a plurality of current students;

generating granular assessment data, the generating including;

processing assessment documents by capturing and evaluating each free-form mark provided in response to at least one assessment item administered to the current students;

accessing the granular assessment data and at least one attribute including one of student data, educator data, and educational material data for each of the current students and a plurality of predecessor students;

for a first one of the current and predecessor students, processing a clustering algorithm using the granular assessment data and the at least one student, educator, and educational material data;

using a mapping algorithm in accordance with the clustering algorithm, mapping a second one of the current and predecessor students to the first one of the current and predecessor students;

using the mapping, correlating one of each current student to a course of action associated with a cluster of predecessor students and a cluster of current students to a course of action associated with a predecessor student; and

,

for the cluster of current students and current student, outputting a course of action associated with the one of the predecessor student and cluster of predecessor students, respectively;

wherein each student of the plurality of students is associated with a D-dimensional vector encoding granular assessment data for that student, wherein the vector includes values indicative of an evaluation of each answer to D respective questions, wherein the value for each question of the D questions is +1 for a correct answer and −

1 for an incorrect answer; and

wherein the clustering includes determining a similarity of a first student having an associated vector u and a second student having an associated vector v, and applying the formula 0.5*(1+cos(u, v)) to u and v.

2. The recommender system in accordance with claim 1, wherein the granular assessment data is obtained from at least the free-form markings extracted from assessment answer sheets scanned in via a multifunctional device (MFD), wherein each assessment answer sheet is used for recording answers to questions associated with an assessment administered to a student of the plurality of students.

3. The recommender system in accordance with claim 1, wherein the course of action in the request is to make classroom assignments for the plurality of current students.

4. The recommender system in accordance with claim 1, wherein the course of action in the request is to form the plurality of current students into individual groups for performing an educational activity.

5. The recommender system in accordance with claim 1, wherein the request includes a plurality of constraints that control the clustering.

6. The recommender system in accordance with claim 5, wherein the constraints include a plurality of logistical constraints which must be satisfied by the clustering.

7. The recommender system in accordance with claim 5, wherein a clustering algorithm is a constraints based algorithm for satisfying a constraint satisfaction problem.

8. The recommender system in accordance with claim 6, wherein the respective constraints of the plurality of constraints have associated weights, and the clustering is further based on the weights of the associated constraints.

9. The recommender system in accordance with claim 8, the processor further executing programmable instructions for receiving adjustments from the user to at least one of the plurality of constraints and the associated weights;

and repeating the clustering, outputting and receiving adjustments until the user is satisfied with the output results.

10. The recommender system in accordance with claim 1, wherein the data storage facility further stores resource data including attributes related to a plurality of educational resources, the course of action in the request includes recommending an educational resource for at least one of the clusters, and the processor further executes programmable instructions for:

accessing the resource data; and

based on its related attributes recommending an educational resource for respective clusters of the selected number of clusters based on at least one attribute related to the recommended educational resource.

11. The recommender system in accordance with claim 10, wherein the educational resource includes at least one of an educator, physical educational material, digital educational material and an instructional method.

12. The recommender system in accordance with claim 10, wherein the student data includes data indicating needs or preferences of the respective students and the resource data includes data indicating strengths or capabilities of the respective resources, and the educational resource recommended for a particular cluster is recommended when resource data associated with the resource indicates that the resource'"'"'s strengths or capabilities are compatible with needs or preferences of the particular cluster indicated by the student data associated with the students included in the particular cluster.

13. The recommender system in accordance with claim 10, wherein recommending the educational resource is further based on a selected level of performance as indicated by the granular assessment data associated with at least one predecessor student of the plurality of students, which are students of the plurality of students, other than the current students, that have prior experience with a course of action related to the course of action in the request, wherein the granular assessment data that the recommendation is based on is related to an assessment reflective of a student'"'"'s performance based on experience with the recommended educational resource or an educational resource that has attributes which are similar to the attributes of the recommended educational resource.

14. The recommender system in accordance with claim 10, wherein the educational resource includes natural language processing (NLP) reading material which is digitally modifiable to provide a reading level compatible with a reading level associated with the cluster it is recommended for, wherein the reading level associated with the cluster is determined based on the granular assessment data.

15. The recommender system in accordance with claim 1, wherein the clustering is further based on a selected level of performance as indicated by the granular assessment data associated with at least one predecessor student of the plurality of students, which are students of the plurality of students, other than the current students, that have prior experience with a course of action related to the course of action in the request.

16. The recommender system in accordance with claim 15, wherein the student data further includes, associated with each student of the plurality of students, student attribute data describing at least one attribute of the student, and wherein the at least one predecessor student is selected based on the similarity of the student data associated with the at least one predecessor student relative to the student data associated with the plurality of current students.

17. The recommender system in accordance with claim 16, wherein the student data associated with each respective student defines a location within a multidimensional space whose dimensions are each related to a respective attribute, wherein selected attributes of the at least one attribute and selected granular assessment student data are each associated with a respective dimension of the multidimensional space, wherein the selection of the at least one predecessor student based on the similarity of the student data includes selecting based on a relationship between the locations associated with respective students of the at least one predecessor student and respective current students of the plurality of current students.

18. The recommender system in accordance with claim 1, wherein the clustering includes using non-negative matrix factorization.

19. The recommender system in accordance with claim 1, wherein the clustering includes using probabilistic latent semantic analysis.

20. A method for recommending clustering of students into educational groups, the method comprising:

receiving a request at a processor to recommend a course of action related to a plurality of current students;

using the processor, acquiring granular assessment data-including evaluating each free-form mark captured in response to at least one assessment item administered to the current students;

using the processor, accessing the granular assessment data and at least one attribute including one of student data, educator data, and educational material data for each of the current students and a plurality of predecessor students;

for a first one of the current and predecessor students, processing a clustering algorithm using the granular assessment data and the attribute;

using a mapping algorithm in accordance with the clustering algorithm, mapping a second one of the current and predecessor students to the first one of the current and predecessor students;

using the mapping, correlating one of each current student to a course of action associated with a cluster of predecessor students and a cluster of current students to a course of action associated with a predecessor student cluster based on similarity; and

,

for the cluster of current students and current student, outputting a course of action associated with the one of the predecessor student and cluster of predecessor students, respectively;

wherein each student of the plurality of students is associated with a D-dimensional vector encoding granular assessment data for that student, wherein the vector includes values indicative of an evaluation of each answer to D respective questions, wherein the value for each question of the D questions is +1 for a correct answer and −

1 for an incorrect answer; and

wherein the clustering includes determining a similarity of a first student having an associated vector u and a second student having an associated vector v, and applying the formula 0.5*(1+cos(u, v)) to u and v.

21. The method according to claim 20, wherein the clustering is performed in accordance with at least one user adjustable clustering constraint, wherein the at least one adjustable clustering constraint is selected from the group of clustering constraints consisting of:

a student attribute constraint specifying attribute data included in the student data describing attributes related to respective students of the plurality of students to use for determining similarity when the clustering is based on similarity, a differentiation constraint for specifying a degree of differentiation between clusters, a cluster size constraint specifying a minimum number of students to include in each cluster, a maximum number of students to include in each cluster, the selected number of clusters, and an algorithm constraint specifying a clustering algorithm to apply.

22. The method according to claim 20, further comprising:

accessing resource data in the computer database describing attributes related to a plurality of educational resources;

recommending an educational resource for respective clusters of the clusters based on the resource data and the student data corresponding to students included in the respective cluster.

23. The method according to claim 22, wherein the student data includes data indicating needs or preferences of the respective students and the resource data includes data indicating strengths or capabilities of the respective resources, and the educational resource recommended for a particular cluster is recommended when resource data associated with the resource indicates that the resource'"'"'s strengths or capabilities are compatible with needs or preferences of the particular cluster indicated by the student data associated with the students included in the particular cluster.

24. The method according to claim 20, wherein the clustering is further based on a selected level of performance as indicated by the granular assessment data associated with at least one predecessor student of the plurality of students, which are students of the plurality of students, other than the current students, that have prior experience with a course of action related to the course of action in the request.

25. The method according to claim 24, wherein the student data further includes, associated with each student of the plurality of students, student attribute data describing at least one attribute of the student, and wherein the at least one predecessor student is selected based on the similarity of the student data associated with the at least one predecessor student relative to the student data associated with the plurality of current students.

26. The method in accordance with claim 25, wherein the student data associated with each respective student defines a location within a multidimensional space whose dimensions are each related with a respective attribute, wherein selected attributes of the at least one attribute and selected granular assessment student data are each associated with a respective dimension of the multidimensional space, wherein the selection of the at least one predecessor student based on the similarity of the student data includes selecting based on a relationship between the locations associated with respective students of the at least one predecessor student and respective current students of the plurality of current students.

27. The method in accordance with claim 20, wherein the clustering includes using non-negative matrix factorization.

28. The method in accordance with claim 20, wherein the clustering includes using probabilistic latent semantic analysis.

29. A non-transitory computer-readable medium storing a series of programmable instructions configured for execution by at least one processor for recommending clustering of students into educational groups comprising the steps of:

receiving a request to recommend a course of action related to a plurality of current students;

acquiring granular assessment data including evaluating each free-form mark captured in response to at least one assessment item administered to the current students;

accessing the granular assessment data and at least one attribute including one of student, educator, and educational material data that corresponds to each of the current students and a plurality of predecessor students;

for a first one of the current and predecessor students, processing a clustering algorithm using the granular assessment data and the attribute;

using a mapping algorithm in accordance with the clustering algorithm, mapping a second one of the current and predecessor students to the first one of the current and predecessor students;

using the mapping, correlating one of each current student to a course of action associated with a cluster of predecessor students and a cluster of current students to a course of action associated with a predecessor student cluster based on similarity; and

,

for the cluster of current students and current student, outputting a course of action associated with the one of the predecessor student and cluster of predecessor students, respectively;

wherein each student of the plurality of students is associated with a D-dimensional vector encoding granular assessment data for that student, wherein the vector includes values indicative of an evaluation of each answer to D respective questions, wherein the value for each question of the D questions is +1 for a correct answer and −

1 for an incorrect answer; and

wherein the clustering includes determining a similarity of a first student having an associated vector u and a second student having an associated vector v, and applying the formula 0.5*(1+cos(u, v)) to u and v.

30. The non-transitory computer-readable medium according to claim 29, wherein the course of action in the request is to make classroom assignments for the plurality of current students.

31. The non-transitory computer-readable medium according to claim 29, wherein the course of action in the request is to form the plurality of current students into individual groups for performing an educational activity.