This article investigates the relationship between news title's attitude towards China and click/relay rate moderated by new title's attitude towards American, applying quantitative research method of moderation regression. The whole content analysis procedure has two stages. The first is manual content analysis of 10thousand pieces of news. Second, the next step applied the machine learning technology, to reach the left 40thousand pieces of news. This method combines the traditional content analysis with advanced technology, reaching the bigdata for investigation. This article attempts to figure more about the current Chinese public opinion towards America through these data, and then researches the following communication process with selective spiral theory.%本文旨在结合选择性螺旋理论,通过分析网络新闻的传播路径来探究当前中国民众的美国观,主要通过内容分析以及调节回归的量化研究方法来进行研究,其中自变量为"对于中国而言的媒体导向",因变量为"点击量/转发量",调节变量为"对于美国而言的媒体导向".研究方法将传统的内容分析与计算机科学技术结合,分两个步骤:(1)通过人工内容分析方法对1万条新闻标题进行标识;(2)利用机器学习技术标识剩下约3.7万条新闻标题.
展开▼
