Linguistic and non-linguistic elements in detecting (Hungarian) fake news (abstract)
2017. április 19. szerda, 18:17

"HEADS UP!!! REALLY DANGEROUS TEA IS SOLD RIGHT OVER THE COUNTER! IT'S VERY POPULAR, YOU PROBABLY ALSO HAVE A PACKET IN THE KITCHEN BUT YOU SHOULD NEVER EVEN TRY IT! Throw it out or bring it back to the store immediately if you have bought one! It's the most popular tea on supermarket shelves! Here's the full heads-up message >>>>>"

This is an article headline from a Hungarian website, an obvious example of clickbait. The exaggerations and the uppercase letters used in the text are (or should be) clear signs of the deceptive nature of the article, still it was shared by many users on Facebook. What could be the reason for this?

The presentation tries to answer the question by collecting the linguistic and non-linguistic characteristics of fake news. Linguistic characteristics include among others the exaggerating, sensational title, the eye-catching, tabloid-style text, the correct or incorrect use of terms, and style, linguistic quality (the use of all caps, excessive punctuation and spelling mistakes); non-linguistic characteristics are the content; chosen characters; expressive pictures often featuring celebrities, the fake URLs imitating real websites. The corpus was compiled using snowball sampling: manipulative news not originating from big news portals were collected from the social networking website Facebook. This time political and lifestyle type news were not distinguished. Concealed sites (i.e. sites containing news of better linguistic quality reached via non-suspicious links) can be problematic when selecting articles for the corpus, and this study also ignored the potential fake news of otherwise as trusted listed news portals ("alternative facts").

The aim of the study is, on the one hand, to identify the characteristics of Hungarian fake news in comparison to English ones, and on the other hand, to elaborate a system of aspects which help identify fake news.