Loading...

Media Bias Analysis for Persian Text News

Abbaszadeh Hojedki, Mohaddese | 2017

622 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 49529 (31)
  4. University: Sharif University of Technology
  5. Department: Languages and Linguistics Center
  6. Advisor(s): Bahrani, Mohammad
  7. Abstract:
  8. There are different types of media bias. The aim of this study is to analyze media bias by considering two types of it: selection (or coverage) bias and language bias. Thus we have collected some specific news stories or articles -which contain “Iran” as a keyword- from the websites of four news broadcasters that are Al Arabiya, Deutsche Welle (DW), Radio France Internationale (RFI) and SPUTNIK, to build text datasets. For the purpose of comparing and analyzing media bias, the news had to be gathered during two time frames before and after the day the P5+1, European Union and Iran reached Joint Comprehensive Plan of Action (JCPOA). Finally, the collected corpora have amounted to 784 news stories and more than 376000 words. To identify the bias in story selection and to find how the broadcasters have covered them, we have used an unsupervised probabilistic model called “Latent Dirichlet Allocation” for topic modeling. So we can extract word clusters or “topics” and approximate the percentage of each topic's weight. The topic models indicate that the broadcasters have selected different stories for publication more; common stories between broadcasters also have had different weights that represent the non-identical distribution of these stories in the news corpora. To detect language bias, we have built a lexicon of bias Persian words then we have utilized it and some other features for training a supervised model to make prediction for bias-including words in text corpora. The Precision performance measure for this classifier model is 76% and Recall is 77%. Interpreting the results of bias ratio for each broadcaster’s news corpus reveals that using of bias language has been increased in after JCPOA time period in comparison with before JCPOA time period for all four broadcasters. The relative growth rates of language bias are also 10.43%, 9.76%, 5.23% and 3.27% for Al Arabiya, RFI, DW and SPUTNIK respectively
  9. Keywords:
  10. Latent Dirrichlet Allocation (LDA) ; Topic Modeling ; Persian Texts ; Media Bias ; Selection Bias ; Language Bias

 Digital Object List

 Bookmark

No TOC