The College of New Jersey Computer Science Department

The following projects will partially fulfill the practicum experiences that your program of study requires. Two practicum experiences are required to fulfill graduation requirements. We recommend one experience during the junior year and one during the senior year. You may choose any combination of mentored research or internship. Course numbers have changed. Mentored research is either CSC 498 or 499, depending on level of experience. CSC 499 is reserved for when you are continuing work on a project. Internship is now CSC 399. (If you are interested in internship opportunities, please logon to LionsPro through Career Services and contact Dr. Komagata, our internship coordinator. Check the internship listing for more information.)

Please read through the following project descriptions. If you are interested in some of them, schedule an appointment with the faculty member indicated to discuss your interest, qualifications, and schedule. When you are invited to join the project by the faculty member, please fill out the mentored research contract, including signatures. The contract is to be turned in for review by the Department Chair, and then after final approval, you will need to bring a copy of the contract to Records & Registration for inperson registration for that one course.

Research projects are listed in alphabetical order by faculty member’s last name. As new proposals arrive, this list will be updated. If you have a great idea for a project that you don’t see listed, please visit with the faculty member most closely interested in that area and propose a discussion!

Dr. Peter DePasquale:

Web security and PHP enhancements to the department web site - Explore PHP in-depth PHP build building additional functionality to our department web site. Also, help secure the site further by researching about HTML injection and SQL injection and apply what you have learned to our site. Another possible avenue to pursue would be to begin migrating the site to JSPs and Java Server Faces.
Web Filtering – Continue the design, development, and deployment (hopefully) of a community-based web filtering technology. The technology uses Bayesian Statistical Filtering (similar to that used in junk email filtering) to filter web content. Prior experience with PHP, Java and databases are a must.

Mentored research will include regular weekly meetings, directed readings and discussions, as well as development and implementation of various related software components. Solid programming background is required, while familiarity with PHP/JSPs/Java is a plus. For more information on working with me, please refer to my web site. Thanks.

Dr. Deborah Knox:

Linux Cluster Lab: Develop kernel applications to customize or enhance a Linux installation. Learn via systems administration of a six node Linux cluster, suitable for distributed and parallel computing projects. Prerequisites: 325, 330, 340
Forensic Computing: Are you interested in areas of forensic computing such as the preservation, identification, extraction, and documentation of digital evidence? This project will introduce the area through readings and emphasize learning to gather, preserve, and analyze evidence such as deleted and hidden files, encrypted information, illegal software, log files, etc. Help investigate the state of the art and identify necessary support structures. Open source software will be a component of this study. Prerequisites: 325, 330, 340
Other Topics: What’s on your wish list? Please visit with your idea sketched out and we’ll discuss your interests!

Dr. Nobo Komagata:

Dr. Komagata will be on leave during the 2005-2006 academic year and thus will not be offering any mentored research projects.

Dr. Jikai Li:

Network Performance Changes: This research is a continued effort to investigate how a new transport layer protocol can work efficiently and fairly with emerging high-speed wired and wireless network. Last several years have seen the bandwidth of wired network and wireless network increased dramatically. This bandwidth boom has already displayed a significant effect on our daily life and the world. However, due to the limitation of current transport layer protocols, the potential of these high-speed networks is not fully utilized. In the last several years, researchers have intensively investigated the new protocols that can transmit data efficiently and fairly. In this work, we will continue our research to develop new transport layer protocol for high-speed wired and wireless network. This work will use NS-2 to study the performance of network. Although programming is not focus of this research, moderate programming is a must-be. Several questions should be answered at the end of the semester.
- How throughput and goodput are affected by the protocol?
- How is the fairness performance of the new protocol?
The student interested in this research should have C/C++ programming experience, have Networking experience (took networking course) before. At the end of the semester, student is expected to turn in a research report.

If you have a plan to program OpenGL, bring in your plan. I would like to talk with you. You are expected to have previous OpenGL experience and a strong motivation to do something fun.

Dr. Mike Martinovic:

QASRTIIR: The focus of my current efforts is on the theory and practice of design, development, testing and evaluation of hybrid Question Answering systems that combine statistical and linguistic techniques in order to optimize their performance. The system being envisaged and developed (QASTIIR http://www.tcnj.edu/~mmmartin/QASTIIR.html) is an integrated, on-demand, dynamic, modular and flexible system with portable linguistic components that can be moved within the system dynamically. These components can be included ("plugged-in") or excluded ("pulled-out") from processing based on the query and user characterizations obtained during a preprocessing phase. Within the above framework, a future work on the following components of the Question Answering system are envisaged:
1. Named Entity Recognition - design and development of a module that is capable of recognizing named entities within a text (i.e. person names, organization names, locations, monetary epxressions, etc., etc.). Such a system has an important role within the information extraction phase of QASTIIR.
2. Reference Resolution - design and development of a module that is capable of resolving pronomial and definite referencing within a text. Within this framework, a research will be conducted into (ii-a) electronic lexicons and (ii-b) natural language parsers and part-of-speech taggers.
3. Word Conflation - design and development of a module that is capable of doing lemmatization and stemmi ng by always producing minimal stems.

Mentored research will include regular weekly meetings, directed readings and discussions, as well as development and implementation of various related software components for QASTIIR. Solid programming background is required, while familiarity with Natural Language Processing techniques and Artifical Intelligence is a plus. Fall semester enrollment for my mentored research is limited to three seats.

Dr. Norman Neff:

Dr. Neff is not offering any mentored research projects in the Fall of 2005.

Dr. Monisha Pulimood:

Support for Applications on the Grid: Grid Technology is fast revolutionizing business models and scientific research methods. A grid computing system is a distributed parallel collection of computers that enables the sharing, selection and aggregation of resources. This sharing is based on the resources' availability, capability, performance, cost and ability to meet quality-of-service requirements. Bioinformatics, oil and gas exploration, automotive and aerospace engineering, and financial services industries were among the early corporate adopters. Today the Grid spans computers at several universities, commercial and government research facilities.
My research interest is in languages that can produce strongly mobile programs to effectively exploit the Grid. A strongly mobile program (mobile computation) is one that can commence execution at one site, discover a need for a resource at a different site and then halt execution, migrate, and continue execution at the new site. In order to continue (without being forced to start over) the current state of execution must migrate along with the mobile computation. Research activites include researching and setting up infrastructure for an experimental Grid Computing environment; developing and executing performance evaluation studies for mobile computations (programs); implementing modules that provide support for strong mobility in the language and virtual machine.

Dr. Pulimood's mentored research section is currently filled.

Dr. Ursula Wolz:

Ursula Wolz is offering opportunities in Fall 2005 for research on two projects: WHAT, rMusic and Hopewell. More detailed information about the background of these projects can be found at her website: http://www.tcnj.edu/~wolz. Each project has its own site, and there is a summary of research page under construction. Please note that working with Dr. Wolz entails identifying a research interest and finding a "good fit" within her research agenda. You will not be given a detailed assignment, but rather will be expected to take the initiative to develop a research goal and implementation plan in collaboration with Dr. Wolz and possibly other students.

rMUSIC: This project was conceived in collaboration with Mike Massimi who is about to embark on a graduate career in Human Computer Interfaces. rMUSIC is an exploration of collaboration technology and how “natural credibility based voting” can influence the use of highly collaborative computing. A complete system is about to be tested. Work on this project next year will primarily involve porting the underlying architecture to domains other than music, perhaps to a collaborative game environment.

WHAT: Web Host Access Tools is a eight year long project in collaboration with Dr. Liilian Cassel of Villanova University. The goal of the project is to create highly personalized assistance for web search. Research topics include
1. how to construct robust queries (e.g. using natural language, or icon-based methods of query construction),
2. provide informative responses (e.g. better ranking of returned URLS, alternatives to single ranked list displays.),
3. how to develop, update and maintain a robust personalized context (e.g. through information retrieval and data mining techniques that may or may not be combined with machine learning techniques such as Support Vector Machines and Genetic Algorithms.
Hopewell: Begun in 1999, this is a recognized community outreach project. In collaboration with members of the Hopewell Valley Regional School District we are building collaboration technology that allows groups of teachers to address curriculum issues in a dynamic, organic fashion. Through "wizard of oz" support by TCNJ students a robust "map" of the K-12 curriculum and related resources and materials is being developed. This is not intended as a repository, but rather is viewed as a rapidly changing collaboratively developed artifact. A major concern of this project is maintaining privacy and safety for the constituents. Research issues range from database design and administration to knowledge representation, to real-time collaboration software to data mining to interface design. (The Clay project which is current dormant, is vehicle through which real-time collaboration on an artifact can be explored.) If you are at all interested in user interface issues, including natural language processing, portal design, or knowledge extraction, this project can offer you venues for exploration.