Sparse Factor Analysis for Learning Analytics and Content Analytics

Abstract

A mechanism for facilitating personalized learning. A computer receives graded response data including grades that have been assigned to answers provided by learners in response to a set of questions. Output data is computed based on the graded response data using a latent factor model. The output data includes at least: an association matrix that defines a set of K concepts implicit in the set of questions, wherein K is smaller than the number of questions, wherein, for each of the K concepts, the association matrix defines the concept by specifying strengths of association between the concept and the questions; and a learner knowledge matrix including, for each learner and each of the K concepts, an extent of the learner's knowledge of the concept. The computer may display a visual representation of the association strengths in the association matrix and / or the extents in the learner knowledge matrix.

Description

PRIORITY CLAIM DATA[0001]This application claims the benefit of priority to U.S. Provisional Application No. 61 / 790,727, filed on Mar. 15, 2013, entitled “SPARSE Factor Analysis for Learning Analytics and Content Analytics”, invented by Richard G. Baraniuk, Andrew S. Lan, Christoph E. Studer, and Andrew E. Waters, which is hereby incorporated by reference in its entirety as though fully and completely set forth herein.STATEMENT OF GOVERNMENT RIGHTS[0002]This invention was made with government support under NSF Grant No. IIS-1124535 awarded by the National Science Foundation, Office of Naval Research Grant No. N00014-10-1-0989 awarded by the U.S. Department of Defense, and Air Force Office of Scientific Research Grant No. FA9550-09-1-0432 also awarded by the U.S. Department of Defense. The government has certain rights in the invention.FIELD OF THE INVENTION[0003]The present invention relates to the field of machine learning, and more particularly, to mechanisms for: (a) exposing the...

AI Technical Summary

Benefits of technology

This patent describes a method for creating a personalized learning experience for users based on their graded response data and preference values for content items. The system uses a latent factor model to compute an association matrix and a concept-preference matrix, which can be displayed to the user in a visual representation. The technical effect of this innovation is a more targeted and effective learning experience for users, based on their unique needs and learning styles. The content items can be provided by entities like businesses, government agencies, or educational institutions and can be accessed online.

Problems solved by technology

Requiring domain experts to label the questions with tags is an obvious limitation to the approach, since such tags are often incomplete or inaccurate and thus provide insufficient or unreliable information.

Images