CIRI-CIRI LINGUISTIK DI ARTIKEL BERITA HOAKS DALAM PEMILU AS 2016: ANALISIS SEMANTIK = LINGUISTIC FEATURES IN FAKE NEWS DURING 2016 U.S. PRESIDENTIAL ELECTION: A SEMANTIC ANALYSIS

This study focuses on (1) analyzing and manually extracting linguistic features of the text within sentences of fake news articles found on various digital media platforms by utilizing an analysis method used by experts such as Hezili (2010) and F. Hasan (2018), and (2) overseeing the distinct linguistic differences between fake news stories and genuine news stories. This study uses qualitative descriptive research method, and semantic analysis with text extraction method. The data was obtained from a popular fact-checking website (Politifact.org). The analysis was conducted in the following steps; 1) The writer obtains fake news articles from a fact-checking website, 2) the writer then analyzes and extracts the linguistic features of selected fake news articles as the main data, 3) the writer categorized the selected features into five different parts of linguistic features, and 4) the writer determines the results from the frequently used linguistic features in fake news articles. Linguistic features are divided into five different parts; 1) Lexical features, 2) orthographic features, 3) grammatical features, 4) discourse features, and (5) graphic features. Based on the analysis results, the writer concluded that three selected fake news articles misused some linguistic features compared with genuine news articles to manipulate the narrative they present and mislead the readers, deeming it as true; emotionally charged diction, sensational interjections, non-standard capitalization, frequent use of passive voice, fragmented coherence, short paragraph structures, and manipulated relevance.

Keyword : Linguistic features, fake news, U.S. Election