The central theme in the course is the role of data management in the realisation of search engines, and specifically those for multimedia data.
The course is part of the Media and Knowledge Engineering Master's program (4 ECTS). It is organised as a series of seminars, with emphasis on active participation. After two introductory lectures, the remaining knowledge is acquired by studying recent research publications. The course's format is an ideal preparation for doing your MSc thesis work.
The course consists of two parts: a series of seminars and a practical assigment.
For the seminars, each student is assigned one or two papers (depending on length of paper). The papers are presented by these students at class (about 20-25 minutes), and then discussed in the group (about 15 minutes).
The idea of the practical assignment is to select one of the research topics discussed in the course, and make a small prototype using MonetDB-XQuery and/or its PF/Tijah information retrieval extension, to illustrate the core research idea in some practical setting; e.g., you could make an image search system for your digital photos, a database-driven photo website, perform an experiment in video classification, etc. The primary goal of the assignment is to get some `hands-on' experience with open source software, and the prototyping of research ideas. But we will no means stop you from creating a fancy demo system of course!
The grade is based on active participation in class, the quality of the presentation, and the result of the practical assignment.
| When | What | Pres. 1 | Pres. 2 |
|---|---|---|---|
| April 7th | Introduction, Database Architecture | - | - |
| April 21st | XQuery and XML Information Retrieval | - | - |
| May 12th | Dataspaces, Entity Ranking | - | - |
| May 19th | Inverted files, Top-K processing | - | - |
| 22/5 | Exploiting user interaction data | - | - |
Lectures take place in CT-D (moved from CT-F), Tuesday mornings, 10:45-12:30.