媒體對同一主題的報導經常有明顯差異,本文以報紙報導風格為研究目標,比較臺灣四大報《蘋果日報》、《自由時報》、《聯合報》、《中國時報》的報導風格。避免報紙立場及題材造成的干擾,本文選取每天頭版頭條新聞,資料期間為2012年至2018年,其中頭條標題涵蓋四大報,但頭條內文僅有《蘋果日報》、《自由時報》。本文分析常用詞彙、句長、標點、虛字,發現四大報頭條標題及內文的用字風格不完全相同,其中標題的Jaccard指標將同時期報導放在同一群,Yue指標將四大報分成三群;內文分析發現《自由時報》傾向政治新聞,《蘋果日報》傾向社會新聞,以能夠代表兩報的常見詞彙為解釋變數,統計及機器學習模型都能準確地區隔《自由時報》或《蘋果日報》。
Like an author's writing style, every newspaper has its own opinion and narrative methods. In this study, our goal is to explore the reporting styles of Taiwan's four major newspapers (Apple Daily, Liberty Times, United Daily News, and China Times) and to compare their differences. We chose the headline news for analysis to prevent the influence of nuisance factors, such as differences in political positions and target audience. The newspaper headlines considered are between 2012 and 2018. The titles of headlines are considered for all four newspapers but the content of headlines is available only for Apple Daily and Liberty Times. We first applied the methods of exploratory data analysis (EDA), such as Jaccard index and Yue index, for the word frequencies and word types to evaluate the similarities between four newspapers. In addition, we also compare the length of sentences, punctuation, and function words of four newspapers. Finally, we used statistical and machine learning methods to distinguish if the news articles are from Apple Daily or Liberty Times. The analysis results showed that there are significant differences between four newspapers. For the headline titles, the Jaccard index grouped titles by time and the Yue index grouped titles by the media (i.e., three groups). For the headline contents, we found that the articles of Liberty Times emphasize political terms and those of Apple Daily focus on social affairs and crime problems. All classification methods have high accuracy if we chose frequently used two-word phrases as the independent variables.