Evaluating event detection on citizen journalism website using tagging : a case study of iReport

Increasingly individuals are turning to online sources for their daily news and in recent years citizen journalism has emerged, competing with mainstream media for the latest breaking information. Citizen journalism has gained momentum with recent technological developments in mobile devices and...

Full description

Saved in:
Bibliographic Details
Main Author: Jeeha Anuksha
Other Authors: Goh Hoe Lian, Dion
Format: Theses and Dissertations
Language:English
Published: 2009
Subjects:
Online Access:http://hdl.handle.net/10356/18233
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Increasingly individuals are turning to online sources for their daily news and in recent years citizen journalism has emerged, competing with mainstream media for the latest breaking information. Citizen journalism has gained momentum with recent technological developments in mobile devices and high speed of internet access. Citizen carrying a digital camera are more likely to capture first hand footage of major new breaking events as compared to professional journalists. This dissertation reports the results of the study to investigate user activity and the type of news reported by citizens by analyzing the tags that users attached to the videos or pictures they upload on iReport, a CNN initiative on citizen journalism. The dataset was taken from iReport during the month of April to June 2008. The tag metrics that have been used during this study include tag quantity, tag growth, tag reuse and tag entropy. The study also compares whether citizen journalism is leading or lagging behind mainstream news by comparing the relative frequency of tags on a time scale. It was discovered that out of 10 events identified from mainstream news, citizens have reported 6 of these events on iReport. Moreover, it was found that iReport was more popular with users from US such that various events related to US were detected using the event detection methodology.