Get news from the Web and analyze them!

ralfswp08.pdf

CANCELED

REASON: The organizer (Ralf Krestel) did not want to assign all the students who selected this topic as first choice.

Ernesto diaz@l3s.de

CANCELED

REASON: The organizer (Ralf Krestel) did not want to assign all the students who selected this topic as first choice.

Ernesto diaz@l3s.de

CANCELED. No students selected this topic.

Ernesto diaz@l3s.de

Everyone writes blogs these days, but all the content is spread over different blog sites and it is nearly impossible to find useful information. Current blog browsers suffer from one main problem – You never know how information across different blog sites are related or differ; or how it the ideas and topics grow or fade out. What we need is …. to build a Better Blog-o-Sphere Browser … Find out more!

Objective: Design and Develop a Blog Analysis and Browsing Tool

Recommended Team size : 3 - 6

Cool Stuff to Play With: XML, Java, C, MySQL; Web 2.0, Data Mining, Natural Language Processing, Information Visualization and of course ….. Blogs

Contact: stewart@L3S.de

softwareproject_ecosystem.ppt

CANCELED

Objective: Develop a tool for the visualization of the functionality of the inverted index. We will visualize the insert, update, delete and search of the documents as well as the ranking algorithms. We will consider privacy and confidentiality issues.

Recommended Team size: 4-8 (We will adjust the complexity of the task to the group size).

Tags: Java, Databases, Inverted Index, Information Retrieval, Security, Confidentiality, Zerber, Ranking, SWING, SWT, Encryption, GUI, Algorithm Visualization, Indexing.

invertedindex_visualizer.ppt

References

[1] Manning, C., Raghavan, P., Schütze, H. Introduction to Information Retrieval. Cambridge University Press 2008 (a pdf version of the book is also available on this page)

[2] Zerr,S., Demidova, E., Olmedilla, D., Nejdl, W., Winslett, M., Mitra, S. Zerber: r-Confidential Indexing for Distributed Documents. EDBT 2008 Nantes

[3]Zerr,S., Olmedilla, D., Nejdl, W. Zerber+R: Top-k Retrieval from a Confidential Index. Internal Report

Objective: Develop a method for detecting events from timestamped data points. The method involves the following three main steps: data representation (graph representation); graph cutting based on semantics; and graph cutting based on temporal patterns.

Recommended Team size: 5 - 8 (We will adjust the complexity of the task to the group size).

Requirements: Java, Databases, Basic knowledge and passion for data mining

softwareproject.ppt

Objective: OKKAM Match is a JAVA project which contains various modules that perform the same task. The OM-Benchmark is a tool for reporting the performance & efficiency of the OKKAM Match modules.

Recommended Team size: 3 - 6 (We will adjust the complexity of the task to the group size).

Requirements: Eclipse, Java

om-benchmark.ppt