Computer Science EducationMy interests in computer science education focuses on the implementation and assessment of active learning methods and the development of curricula for CS0 courses and broadening participation in computer science.
Tablets across the Sciences:
Duke University is developing a new model for technical education designed to improve student understanding and interest in course material, better integrate material across courses, provide more realistic depictions of science inquiry, and improve student retention. We seek to evaluate the pedagogical and technical support infrastructure required for successful integration of tablet technology in this model of education. As part of the project, we will develop, adapt, and evaluate curricular materials for implementations in Electrical and Computer Engineering, Computer Science, and Chemistry courses. Along with the pedagogical concerns of working with tablets, we will also address the administrative support infrastructure required. The three implementations will also have different use conditions: continuous use for lab, lecture, and homework, group use in an active lecture, and in a course laboratory as electronic laboratory notebooks. The faculty will produce video modules exhibiting best practices and will work with an external assessment specialist to thoroughly evaluate the effectiveness of the project's goals.
JiTT is a teaching and learning strategy based on the interaction between web-based study assignments and an active learner classroom. The essence of JiTT is the feedback loop formed by the students' preparation outside the classroom that shapes their in-class experience. The goal of JiTT is to use feedback to guide teaching and to empower and motivate learners. This project will produce course materials utilizing the JiTT framework. The course is structured as series of narratives or case studies. These case studies present real examples of computer science applications that engage a number of fundamental computer science topics. First, these case studies will highlight current work from the field. Second, web assignments ask students to model phenomena, complete warm up exercises, and experiment with simulations. Finally, the case study includes a discussion of the philosophical and social questions that arise when computing is applied in the particular area.
Extrospective Computer Science:
Introductory computer science courses traditionally are introspective. The problems and motivating examples are often drawn from domains that primarily interest computer scientists such as data processing (searching, sorting, file manipulations, and so on). The principal investigators form an interdisciplinary team of computer scientists, a biologist, and a physicist who propose that a more outward-looking or extrospective approach to computer science will better motivate, interest, and prepare students from a variety of backgrounds. We propose to build web-based modules that explore computer science concepts in the context of scientific applications outside of computer science. These modules will highlight current research from these outside fields as background; have an experimental section where students write, modify, and test algorithms; and conclude with a discussion on the philosophical and social questions that arise when computing is applied in the particular field.
RoboticsFacilitating intelligent control of mobile robots
I focus on developing a pedagogically sound software architecture for cognitive robotics. By implementing parameterized high-level actions, users can focus their efforts on developing control algorithms. We propose to develop methods for:
RoboCupJunior: Exhibitions of Problem Solving, Teamwork, and Creativity
RoboCupJunior is a project-oriented, team-based academic enrichment program for students at Chewning Middle School in Durham. The objective of the program is to foster interest and competence in science, mathematics, and computers, while developing problem-solving skills, enabling creative thinking and design, and providing a domain for application of scientific concepts. RoboCupJunior is an educational initiative where teams of students build and program mobile robots to perform a variety of challenges. Students will learn and apply the scientific, mathematical, and technological fundamentals behind the construction of robots and the design of control algorithms. Robotics-related research demonstrations at Duke will supplement weekly meetings with student mentors from Duke University. Students will exhibit their work in competitions. The project participants will conduct a workshop for North Carolina teachers to disseminate tested instructional materials in 2007 and hold a RoboCupJunior regional exhibition the following year.
Social Networks & Collaborative FilteringCoBib: Collaborative indexing and annotation of bibliographic databases
This project will build a web repository of research papers where users could input papers and citations, annotate files with comments, and view other users' collective judgments. This database would have significant utility as research aid and as an educational tool. If a student wanted to find out about an area, he or she could see what papers people in that area were reading and quickly be able to gain from the collective wisdom of the peers. This project will extend the work of projects like CiteSEER by enabling collaborative annotation and comments by users in and across research areas. Some of the main problems to be tackled include:
We propose an alternate introduction into computer science by studying the topics that arise from analyzing and modeling social networks. We are developing modules that should capture interest and provide a compelling yet intellectual rich area of study. One module includes using real-world data from sites such as facebook.com to analyze the properties of social networks using graph algorithms and tools. Another example involves using students' music listening behavior with their iPods to predict new musical interests using collaborative filtering algorithms and recommender systems.
Testbed for telemedicine applications in developing countries
In a partnership between the Duke Computer Science department, the CUREs nonprofit business plan competition, and Engineering World Health, the CUREs Telemedicine Project will develop a sustainable nonprofit business based on an innovative handheld computer software solution designed to provide medical diagnosis and treatment to underserved patients in remote communities across the developing world. The CUREs Telemedicine Architecture uses the AJAX framework to develop web applications for Windows mobile clients that will be used on wireless mesh networks. Building upon current research and available toolkits for mesh networking, we will build a testbed for Telemedicine applications designed for remote communities. Despite the lack of widespread high-speed Internet connectivity, computational resources and medical or technical expertise, the project will demonstrate that technology can be well integrated and adapted into the culture of Nicaragua. Domain and assessment experts will assist us in evaluating the technical and pedagogical validity of the program.
|Last updated Fri Jan 09 17:51:46 EST 2015|