For several years now, various machine learning methods have been used to analyze digital media content, including in search engines and automatic translations. But what do these methods offer for quantitative content analysis as applied in the social sciences? This book contrasts the methodological and research-practical peculiarities of automatic content analysis with those of classical manual coding. It then discusses the advantages and disadvantages of machine learning compared to other computer-based text analysis methods. Practically, the potential of this approach is evaluated through an extensive analysis of online news. An experimental study focuses on classification quality on one hand and effectiveness on the other.