Information for topics

Topic Id:

ID topic:	373
Partner Email:	guvenir@cs.bilkent.edu.tr
Project Title:	New Event Detection and Topic Tracking Based on Suffix Tree Concept
Abstract:	Our group aims to develop a system which detects and clusters events from different news sources. It will use data fusion approach to evaluate news stories by clustering them on suffix trees and present the detected events on news portals. The motivation behind this idea is to facilitate users with access to the stories of the news that they are reading. Readers will be able to see automatically detected news stories in chronological order. Although the algorithm that we are proposing is independent of the language, we will use an existing Turkish news test collection and evaluate our algorithm using this dataset. Our plan for this project is: 1. Design and development of the algorithm that will be used to detect and cluster new events from a given set of news stories, 2. Evaluating the algorithm with an existing Turkish news test collection, 3. Integrating our system to Bilkent News Portal of Bilkent Information Retrieval Group and make it available for public use.
Advisor:	Halil Altay Guvenir
Link:
Degree:	Bachelor
Keywords: