Benchmark Dataset for Propaganda Detection in Czech Newspaper Texts

Barrón-Cedeño

2022

Data

In this digital era, people rely on the internet for their news consumption. As people are free to express their opinions on social media, much information shared on the internet is loaded with propaganda. Propagandist contents are intended to influence public opinion. In the mainstream media or prominent news agencies, the authors’ and news agencies’ own bias may impact in the news contents. Hence, it is required to detect such propaganda spread through news articles. Detection and classification of propagandist text require standard, high-quality, annotated datasets. A few datasets are available for propaganda classification. However, these datasets are mostly in English. Hindi is the most spoken language in India, and efforts are needed to detect its propagandist contents. This research work introduces two new datasets: H-Prop and H-Prop-News, which consist of news articles in Hindi annotated as propaganda or non-propaganda. The H-Prop dataset is generated by translating 28,630 news articles from the QProp dataset. The H-Prop-News dataset contains 5500 news articles collected from 32 prominent Hindi news websites. We experiment with the proposed datasets using four supervised machine learning models combined with different feature vectors and word embeddings. Our experiments achieve 87% accuracy using Logistic Regression with TF-IDF feature vectors. The datasets provide high-quality labeled news articles in Hindi and open new avenues for researchers to explore techniques for analyzing and classifying propaganda in Hindi text.

Section: Related Workmentioning

confidence: 99%

H-Prop and H-Prop-News: Computational Propaganda Datasets in Hindi

Barrón-Cedeño

2022

Data

“…The tremendous amount of data generated by social media platforms and the dynamic nature of these platforms makes it hugely difficult to discover, analyze, and predict propaganda content. Most of the research work is focused on either specific event or time duration (Johnston & Weiss, 2018), (Kellner et al, 2019), (Baines & O'Shaughnessy, 2014), (Derbas et al, 2020), (Neyazi, 2020), (Baisa et al, 2019) for propaganda analysis. The enormous volume of data, it's multi-media and multi-lingual nature creates challenges like discovering the trends and hidden patterns, connections between social bots, and techniques used for propaganda.…”

Section: Technical Challengesmentioning

confidence: 99%

A Systematic Comparison of Machine Learning and NLP Techniques to Unveil Propaganda in Social Media

Journal of Information Technology Research

2022

With the technological advancements and its reach Social media has become an essential part of our daily lives. Using social media platforms allows propagandist to spread the propaganda more effortlessly and faster than ever before. Machine learning and Natural language processing applications to solve the problem of propaganda in social media has invited researchers attention in recent years. Several techniques and tools have been proposed to counter propagation of propaganda over social media. This work pursues to analyse the trends in research studies in the recent past which address this issue. Our purpose is to conduct a comprehensive literature review of studies focusing on this area. We perform meta-analysis, categorization, and classification of several existing scholarly articles to increase the understanding of the state-of-the-art in the mentioned field.

“…Trusted, satire, hoax and propaganda 2017 corpus (TSHP-17) (Rashkin et al, 2017) and Hyperpartisan News Dataset from SemEval-2019 (Saleh et al, 2019) are the prominent data sets used for the analysis of news articles. Some studies (Popat et al, 2019;Wang et al, 2018;Qazvinian et al, 2011;Baly et al, 2020;Kwon et al, 2013) have worked in the direction of rumor detection and factchecking, whereas some (Saleh et al, 2019;Barr on-Cedeño et al, 2019;Rashkin et al, 2017;da San Martino et al, 2020;Baisa et al, 2019) have worked to uncover the political propaganda in news articles. In their works, the authors (Joshi et al, 2018;Alhindi et al, 2019;Gupta et al, 2019;Hua, 2019) have explored the various NLP techniques along with neural networks and deep learning models to uncover fine-grained and sentence level propaganda in news articles.…”

Section: Propaganda Through Biased Newsmentioning

confidence: 99%

Propaganda analysis in social media: a bibliometric review

2021

IDD

Purpose This paper aims to examine the trends in research studies in the past decade which address the use and analysis of propaganda in social media using natural language processing. The purpose of this study is to conduct a comprehensive bibliometric review of studies focusing on the use, identification and analysis of propaganda in social media. Design/methodology/approach This work investigates and examines the research papers acquired from the Scopus database which has huge number of peer reviewed literature and also provides interfaces to access required for bibliometric study. This paper has covered subject papers from 2010 to early 2020 and using tools such as VOSviewer and Biblioshiny. Findings This bibliometric survey shows that propaganda in social media is more studied in the area of social sciences, and the field of computer science is catching up. The evolution of research for propaganda in social media shows positive trends. This subject is primarily rooted in the social sciences. Also this subject has shown a recent shift in the area of computer science. The keyword analysis shows that the propaganda in social media is being studied in conjunction with issues such as fake news, political astroturfing, terrorism and radicalization. Research limitations/implications The lack of highly cited papers and co-citation analysis implies intermittent contributions by the researchers. Propaganda in social media is becoming a global phenomenon, and ill effects of this are evident in developing countries as well. This denotes a great deal of scope of work for researchers in other countries focusing on their territorial issues. This study was conducted in the confines of data captured from the Scopus database. Hence, it should be noted that some vital publications in recent times could not be included in this study. Originality/value The uniqueness of this work is that a thorough bibliometric analysis of the topic is demonstrated using several forms such as mind map, co-occurrence, co-citations, Sankey plot and topic dendrograms by using bibliometric tools such as VOSviewer and Biblioshiny.