Project

This document describes the mechanisms and modalities of your course project. Listed at the end of this description are several possible project themes. You will form groups of four and chose one of these project areas. The areas will be assigned on a first come, first served basis. In general, I will expect that each group will chose a different project theme. If you chose the same area as another group, then you must make sure that your focussed projects do not overlap. The projects are a mix of those that will involve heavy programming and light experimentation, and those that will involve light programming but detailed experimentation and analysis.

• In the first two weeks, you will do preliminary reading in the area. I suggest you look at recent journal articles, web pages of research projects in this area, technical magazines etc., and narrow the focus of your project. The course resources web page will contain pointers to many ongoing projects in industrial and academic research labs -- this could provide you with a starting point. Other places to explore include IEEE TKDE, ACM TODS, the KDD Journal, proceedings of the annual KDD conference, IEEE ICDE, ACM SIGMOD, ACM DMKD etc.
• You will meet with me and seek approval for your proposed project by 3/6/99. At this meeting, please present a 1-2 page document outlining the project specification. This document should reference the reading you have done thus far, describe deliverables, and include a timeline. It should also have a URL for a web page you will maintain for your project. I suggest that you appoint, by rotation, a person to manage the group. This person can take care of details such as setting up meetings and so on.
• During the week of April 8, you will give 20-25 minute presentations which reflect both the state-of-the-art/practice in the area you are working on, and also gives an outline of your own work thus far.
• Your final project report and demo will be due during the exam week. You will need to arrange for a demo during either the final week, or the week just preceeding it.

1. Mining Multimedia Databases
   1. Generating Hierarchical Visual Summaries from Videos/Image Sequences
   2. Extracting and Mining Characteristics ( and metadata) from Images/Videos
2. Data Warehousing/Mining
1. Comparitive Analysis of Algorithms
2. Characterization Algorithm behaviour in presence of Noise.
3. Web Mining
   1. Mining Web Access Logs by Implementing an Algorithm
   2. Compartitive Analysis of Algorithms
   3. Combining Document Analysis Techniques with Log Mining
4. Web Personalization: What data should be gathered ? How to use the results of mining to deliver personalized content ?
5. Intrusion Detection through Mining Usage patterns.