Project Description for IS698 Term Project

You can choose one topic among any category below for your term project. The purpose of this project list is to provide you with choices of different aspects of database systems, both theoretically and practically. You can work as a group (up to 3 with approval) or individually. If you decide to work as a group, you have to specify the part done by each member of the group both in your proposal and in your term report. Everyone (or a team) must submit a proposal. The proposal should include the title, description, significance, hardware, and software involved, the procedure to work on, and the items to be delivered at the end of the project.

A. Application Development or Implementation

• Multimedia DBMS Implementation: Implement a small multimedia database system in a chosen application. You can use any tool.
• Query Optimization: Experiment with join optimization techniques of Oracle.
• Startup.com: Assume you are creating a start-up company for E-business. Limit the scope and create a sample prototype assuming you are selling authentic CDs(or books, paintings, photos, dolls) you collected.  For simplicity, all the payments are done via credit cards. This could be done in Access 2000 or later version or Oracle.
• Experimentation with ProCube: ProCube is a commercial OLAP tool. Create a sample system and evaluate the interface and OLAP functions. If possible, compare and contrast with other OLAP tools.

B.  Survey or Research Topics

Pick up a topic and produce about 20 (single space) pages long report based on your study. Survey topics need to study the state of the art and the state of practice of a given problem. A survey should be comprehensive, include as many summary tables and figures as possible, and discuss strengths and weaknesses of approaches or systems. Report should be organized in a logical sequence. References should be properly cited and listed at the end of a report. A research topic does an in-depth study of a chosen topic. A research topic is prefixed by symbol ®.

• Multimedia DBMS: Study the state of art and the state of practice in multimedia DBMS. What are hot issues which will be addressed in next few years? Some suggested topics include: multimedia data model, content-based retrieval, or multimedia indexing. You may focus on just one topic.
•  ® Analysis and Experimentation of Class Patterns: I have identified a set of class patterns that can be used for OODB design or OO domain analysis. Apply the class patterns to case studies in several different domains (e.g., medical, GIS, multimedia, CAD/CAM) and analyze the results.
• ® A Comparative Study of DB Design Using UML: There are several books for database design using UML. Perform a comparative study.
• ® A Comparative Study on EM modeling and Class Modeling: What are similarity and differences between ERD and Class diagrams of the UML?  What kinds of classes do not appear in ERDs?
• ® Normalization and Class Modeling: Can we apply normalization rules to class modeling? Can we have class normal form?
• ® Rules for Ternary Relationships: Study ternary relationships and identify and experiment rules for using ternary relationships.
• ® Web Data Extraction: What kinds of useful data can we extract from the web and how? Either perform a survey or a case study.
• ® DB Modeling and Design for Web-based Systems: Study the issues and techniques for database modeling and design for web-based systems and e-commerce (customer, merchandise, billing, shipping, etc), handling of security such as credit card and SSN, return policy and procedures, etc. Other issues could include the handling of multimedia and semi-structured data, translation of paper catalog into a web database, DB design supporting user interface (navigation, hyperlinks, store layout, FAQ, helps, services, policies, shopping cart metaphor, …), schema and data evolution (category of products, naming, prices, …)
• ® A Comparison of Object-Relational DBMS, DB2, Oracle 9i, Informix, and Microsoft SQL Server: Compare and contrast the object-oriented features of popular object-relational database systems. Compare data types and their size and implementation, schema characteristics, SQL variations, indexes and other storage characteristics, locking, optimization characteristics, distributed characteristics, web interface, etc.  The comparison must be done based on the most recent versions.   
• ® Classification of CRM (Customer Relationship Management) Segmentation: Create a taxonomy of customer segmentation in CRM. For each segmentation, identify what measures can we use to find the segmentation, what we can do for each segmentation type, and what kinds of database structural support do we need for each segmentation.
• Use of Data Mining for CRM: Techniques and Limitations: Survey what data mining can do and cannot do for CRM.
• ® A Survey on Medical Information Systems: Survey various DB-related issues in hospital, clinical information systems.
• ® A Survey on Text Mining: Survey various techniques and algorithms of text mining.
• ® Data Mining for Medical Informatics: Survey data mining techniques used in medical informatics area.
• ® Text Mining for Medical Informatics: Survey text mining techniques used in medical informatics area.
• ® A Survey on Cyber Security and Cyberforensics: Survey database issues and technologies for cyberforensics.
• ® Designing MOLAP Data Cubes: How do you design the datacubes for MOLAP (Multidimensional Online Analytical Processing).

C. Your own topic

            If you have other topics under either implementation or research category, submit a proposal, for approval, including nature and scope of the problem, methodology, involved hardware, software, deliverable at the end of project completion, and any known references.