Projects


Education Technology

    K12@USC: The K12@USC project focuses on the design, development and evaluation of technology-based learning resources in math and science. We create intelligent tutoring systems (ITS) to provide middle and high school students with individualized instruction based on prior achievement, cognitive skills, and learner motivation. We have a special focus on creating learning systems to reach students who have traditionally not become highly engaged with math and science.

    Our tutoring systems are designed for classroom integration through alignment with standards, inclusion of assessment and reporting tools for teachers, and strong emphasis on learning outcomes. We work with master teachers to create and review the content and scaffolding. The K12@USC systems can also be used at home, or through after-school and community programs. If you would like more information, please contact project director Carole Beal.

Autonomous Learning Algorithms

    Deep Green / Crystal Ball : Crystal Ball uses continuous fluent graphs to help its user avoid surprises when planning and operating in various domains. In particular, for the Deep Green project, we will be focusing on fluent models of battlefield operations, enabling commanders to improve decision-making in the field by being better able to anticipate possible futures.

    Jean : Jean is an architecture that models developmental learning in an autonomous agent platform. Jean acts and learns in computer simulations, building up a richer understanding of the world her over time as she interacts with her environment.

    Wubble World : Wubble World is a project to design autonomous agents that can learn concepts and language from a live human teacher. By providing a fun, interactive game environment, we are able to tap into large amounts of free human instruction.

    Developmental Chunking : Humans naturally segment the world we perceive into understandable chunks. This project seeks to understand the underlying mechanisms and mathematics that enable this automatic "chunking" of our perceived experience.

© 2006, Learning and Development Center, USC ISI. All rights reserved.