an article by Hend S. Al-Khalifa and Rasha M. Al-Eidan (King Saud University, Riyadh) published in International Journal of Web Information Systems Volume 7 Issue 2 (2011)
Abstract
Purpose
Owing to the large amount of information available on Twitter (a micro blogging service) that is not necessarily true or believable, credibility of news published in such an electronic channel has become an important area for investigation in the field of web credibility. This paper aims to address this issue.
Design/methodology/approach
A system was developed to measure the credibility of news content published in Twitter. The system uses two approaches to assign credibility levels (low, high and average) to each tweet. The first approach is based on the similarity between Twitter posts (tweets) and authentic (i.e. verified) news sources. The second approach is based on the similarity with verified news sources in addition to a set of proposed features.
Findings
The evaluations of the two approaches showed that assigning credibility levels to Twitter tweets for the first approach has a higher precision and recall. Additional experiments showed that the linking feature has its impact on the second approach results.
Research limitations/implications
The proposed system is experimental; thus further experiments are needed to prove these findings.
Originality/value
This paper contributes to the research on web credibility. It is believed to be the first which provides a proposed system to evaluate the credibility of Twitter news content automatically.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment