There are several research opportunities with stipends (pay) for computer science undergraduates in the Duke Computer Science Department for summer 2009. These jobs are with different professors.
If you have questions about a particular job, please contact the professor. BUT NOTE that you must apply by emailing your application to Prof. Rodger. General questions about working in the computer science department in the summer can be sent to Prof. Rodger (rodger AT cs.duke.edu). You can apply to live on Central Campus for the summer.
If interested, please email your application
to Prof. Susan Rodger (rodger@cs.duke.edu) by
February 20, 2009.
(WE WILL CONTINUE TO ACCEPT applications until all jobs are filled.).
Include in your application
This page may continually be updated. Some projects are subject to funding and are noted.
We're looking for several undergraduates to join the JFLAP project! JFLAP is educational software that has evolved over the past 18 years and used in the course CPS 140, and used in over 160 countries! Several Duke undergrads have worked on JFLAP. JFLAP allows one to experiment with concepts related to the foundations of how a compiler works, automata and grammars. Several new features will be developed and added to JFLAP. See the JFLAP page for more info on JFLAP.
Students should have taken CompSci 100. CompSci 108 is helpful, but not required. You do not have to have taken CompSci 140. Many students in the past have worked on JFLAP that have not taken CompSci 140.
Hiring several undergraduates to develop materials for teaching Alice, a 3D virtual worlds environment. Materials developed will be used possibly in CompSci 4, high schools and middle schools. In addition students will aid teachers in 3-weeks of training in Alice to middle school and high school teachers in June (June 15-July 2). There are both full time (10 week positions and shorter 3-5 week positions (May-June)). Please indicate which ones you are interested in.
See the Alice resources at Duke page for more info.
Students should have taken CompSci 4 or higher. It is not required if you know Alice, but if you do, please list.
The increasing complexity of networked computing systems make it hard for users to understand and control these systems. A lot of time and money goes into dealing with unexpected system slowdowns, crashes, and software exploits. The Ques system at Duke tackles this problem using innovative data management and mining techniques. Ques treats a computing system as a rich source of data about system configuration and activity---available as rapid and time-varying data streams---and extracts useful information from this data in order to simplify system administration tasks.
I am looking for one or two undergraduate students to work on interesting projects in Ques and to be part of a system-building effort that is challenging and fun. Feel free to contact me if you have questions or want to see some cool demos of Ques
Of companies that have had a major loss of business data, 43% never reopen, 51% close within two years, and only 6% survive long-term. Major loss of data typically comes from natural or human-induced disasters. This project will build a tool to orchestrate the process of disaster recovery that includes (i) planning the rate at which snapshots of data should be taken, (ii) deciding from which points of the software stack should snapshots be taken (e.g., database, file system, or storage controller?), (iii) validating snapshot integrity, and (iv) identifying and loading the best snapshot for fast system recovery after a disaster.
This project involves both research and programming challenges. Knowledge of database systems is a plus.
The goal of this project is to implement and compare the performance of recently developed algorithms for stochastic scheduling, particularly variants of the multi-armed bandit problem. This problem itself is widely studied and has several applications in sensor networks, vehicle navigation and internet auctions.
Learning objectives: Reading papers, Algorithm design and analysis, Markov Decision Processes, Some auction theory.
Requirements: Discrete math, CPS130, Some programming experience, Reasonable comfort with math.
We have been developing course materials for an alternate introduction into computer science by studying the topics that arise from analyzing and modeling social networks. See Link for more information. Our first module centers around the question, "Can we discover research communities from online faculty CVs?" An undergraduate research assistant will assist in developing this module.
Conducting this research will involve applying techniques from social network analysis, information retrieval, Java programming, web page design, and education. Experience in all of these fields is not necessary, but potential applicants should be willing and able to learn the necessary methods quickly.
We're looking for one student to work on developing support materials for our undergraduate courses. This would primarily focus on the continued development of our web-based grading and scheduling program (which uses php and a database back-end), but may also include development of Eclipse plugins, or developing programming toolkits and libraries to support new versions of our courses. Students should have completed Compsci 108, enjoy programming and thinking, and be curious.
We are looking for a bright and motivated student to help develop FaceBook applications that can be used to improve an online user's credibility. The rapid growth in the number and variety of Internet applications has pushed many societal activities and interactions from the physical world into cyberspace (e.g., online shopping, ticket reservation, chatting, publishing, etc.). Unfortunately, identity and trust, two fundamental concepts that mediate human interactions in the physical world, have not found their way into the online world. The old adage, ``On the Internet, Nobody Knows You're a Dog,'' has never been truer: today's Internet provides little assistance for users and end systems to assess each other's identity and trustworthiness.
Fortunately, online social networks contain a rich structure that embeds abundant identity and trust information between users, groups, and network entities and are growing in popularity. Links between users typically represent certain trust relationships, and a user with a large number of connections to other authentic users is likely to be authentic. In addition, social links within social networks can be annotated by users with levels and types of trust, providing more explicit trust information. For instance, FaceBook users can already label the degree of closeness with their friends using an application such as ``Best Friends." Our goal is to develop new FaceBook applications that can better utilize this rich source of trust and identity information.