This chapter presents computational methods for analyzing emotional appeals in natural language argumentation using psychological affective lexicons and sentiment analysis. Building on earlier parts of the book, it operationalizes the Model of Interactional Pathos in Argumentation (MIPA) by identifying emotion-eliciting words in argumentative structures and modeling audience reactions. Using Polish and English pre-election debates, the chapter demonstrates how lexicon-based methods capture emotional appeals in premise-conclusion pairs and how sentiment analysis can detect audience reactions in real-time social media commentary. It introduces a threefold methodological framework: lexicon-based pathos mining, audience response sentiment analysis, and manual annotation of emotional appeals in online interactions. The results indicate strong correlations between emotional appeals and audience sentiment, including unexpected cross-emotional patterns (e.g., anger appeals that generate sadness). A study of Twitter discussions shows that while negative emotional appeals dominate, positive pathos increases engagement and can soften antagonistic responses. The chapter argues that lexicons and sentiment models, though limited in context sensitivity, offer scalable interdisciplinary tools for the empirical study of rhetorical emotion in discourse. It lays the foundation for subsequent chapters on generative models for pathos mining.